The unknown soldier in citrus plants: polyamines-based defensive mechanisms against biotic and abiotic stresses and their relationship with other stress-associated metabolites

ABSTRACT

Citrus plants are challenged by a broad diversity of abiotic and biotic stresses, which definitely alter their growth, development, and productivity. In order to survive the various stressful conditions, citrus plants relay on multi-layered adaptive strategies, among which is the accumulation of stress-associated metabolites that play vital and complex roles in citrus defensive responses. These metabolites included amino acids, organic acids, fatty acids, phytohormones, polyamines (PAs), and other secondary metabolites. However, the contribution of PAs pathways in citrus defense responses is poorly understood. In this review article, we will discuss the recent metabolic, genetic, and molecular evidence illustrating the potential roles of PAs in citrus defensive responses against biotic and abiotic stressors. We believe that PAs-based defensive role, against biotic and abiotic stress in citrus, is involving the interaction with other stress-associated metabolites, particularly phytohormones. The knowledge gained so far about PAs-based defensive responses in citrus underpins our need for further genetic manipulation of PAs biosynthetic genes to produce transgenic citrus plants with modulated PAs content that may enhance the tolerance of citrus plants against stressful conditions. In addition, it provides valuable information for the potential use of PAs or their synthetic analogs and their emergence as a promising approach to practical applications in citriculture to enhance stress tolerance in citrus plants.

Introduction

In citrus, leaf metabolites play vital and complex roles in the defensive response against biotic and abiotic stress. These metabolites included amino acids,^1–6 organic acids,^1,3,6 fatty acids,^2,6 phytohormones,^7,8 polyamines (PAs)^4,9–12 and other secondary metabolites. However, the contribution of non-proteinogenic amino acids (NPAAs) and PAs pathways in citrus defensive responses are poorly understood.

Polyamines (PAs) are ubiquitous organic compounds that are found in all living kingdoms from Prokaryote to Archaea and Eukaryotes, including plants.^13–16 PAs are low molecular weight, aliphatic, polycationic nitrogen-containing compounds. PAs are containing unsaturated hydrocarbons that are occasionally branched with two or more primary amino groups.^{13,15,17–20} At physiological pH, PAs are positively-charged,²¹ therefore, they can bind to negatively-charged macromolecules, such as nucleic acids (DNA and RNA), acidic phospholipids and numerous types of proteins¹³ to stabilize their structure, particularly under stress conditions. The putrescine, spermidine, and spermine are the major ubiquitously found PAs in higher plants.^22–24

In citrus, PAs are synthesized from _L-arginine through two different routes (well-reviewed recently by Killiny and Nehela, 2020).²⁵ The first route is where _L-arginine is converted by arginaseto ornithine, which is then decarboxylated by ornithine decarboxylase (ODC) to form the diamine putrescine. The second route involving arginine decarboxylation by arginine decarboxylase (ADC) to produce agmatine, which catalyzes toputrescine via some catalytic steps. Once putrescine is produced, it is then converted into the tri-amine spermidine by spermidine synthase (SPDS), with the addition of an aminopropyl moiety donated by decarboxylated S-adenosylmethionine (dcSAM) which produced by the activity of S-adenosyl-methionine decarboxylase (SAMDC; aka AMD1). Subsequently, spermidine is converted into the tetra-amines spermine by the incorporation of another aminopropyl group (also from dcSAM) using spermine synthase (SPMS).

PAs are implicated in several biological and physiological processes for growth and development in both prokaryotes and eukaryotes.^22,26 Furthermore, PAs modulate the plant response to various abiotic and biotic stress, including plant-microbe interactions.²² In citrus, these physiological processes include embryogenesis,^27,28 root system architecture,^29–32 shoot system architecture,^33–35 flowering, fruit set, fruit development, and fruit quality,^10,36-41 and stomatal closure, gas-exchange, photosynthesis and chlorophyll fluorescence.³⁵ Furthermore, PAs might be involved in protecting the citrus plant from stress-associated damages and enhancing stress tolerance against biotic^4,12,42–44 and abiotic stress.^{9–11,34,43,45–52} However, the molecular mechanisms behind these roles remain ambiguous.

Accordingly, in this review article, we discuss the potential roles of PAs in citrus defensive responses with an emphasis on the recent advances in response to abiotic and biotic stress. We are going to discuss the PAs-based defensive responses of citrus against environmental stressors including extreme temperatures (low and high), salinity, heavy metals, drought stress. As well as, we are going to discuss the recent metabolic, genetic, and molecular evidence illustrating the potential PAs-based defensive responses of citrus plants to biotic stress including virus and viroid infection, fungal and oomycetes infection, and bacterial infection, particularly, the vector-borne phytopathogenic bacteria. We believe that the knowledge gained so far about the roles of PAs in citrus responses underpin our understanding of PAs-based defensive responses and provide valuable information for the potential use of PAs or their synthetic analogs a promising approach to practical applications to enhance stress tolerance in citrus plants.

PAs-Based Defensive Mechanisms Against Abiotic Stress in Citrus

Role of PAs in citrus response(s) to low- and high-temperature

Citrus(family Rutaceae) is one of the most widely grown fruit crops worldwide,⁵³ where it grows within 35°-40° north and south latitude.^53,54 Citrus species are considered to be native to the tropical and subtropical geographical areas.^53,54 Therefore, citrus trees are not generally chill hardy. Among the common citrus species, C. reticulata was reported to be the hardiest variety to frost. C. reticulata can tolerate the temperature as low as −10°C for a short period, however, realistically temperature above −2°C is required for successful gardening. Although the low, butnon-freezing temperature, may have some pros such as reducing disease severity, controlling insect pests, and/or modify ripening processes, it may also cause chilling injury (CI).⁵⁵ Low-temperature tolerance in citrus plants is a highly dynamic stress-response phenomenon and has been reported to be associated with complex crosstalk between various biochemical and physiological processes.⁵⁵ Among the biochemical defensive mechanisms to low-temperature stress, changes in PAs and their biosynthetic genes have given considerable attention as a key player against chilling because of their anti-senescence activity and antioxidant- and free-radical scavenging properties.^56–58

Although many previous studies evaluated the changes in the endogenous content of major PAs compounds in citrus and its relative species as affected by low- and/or high-temperature stress (Table 1), very few studies were carried out on leaves tissue,^9–11 while most of these studies were done on fruits.^59–67 The data presented in Table 1 demonstrate a positivecorrelation between putrescine content and citrus tolerance to low- and/or high-temperature. Generally, putrescine was increased in most cases, except in the studies of Martínez-Romero et al. 1999 and Fu et al. 2016.^11,65 For instance, putrescine was decreased when lemon (C. limon cv. Verna) fruits were stored at 2°C for 20 days after treatment with putrescine, calcium, or distilled water.⁶⁵ Putrescine was decreased also when sweet orange (C. sinensis cv. Jincheng) seedlings were placed at high-temperature stress (45°C) for 3 hours.¹¹

Table 1.

Changes in the endogenous content of major PAs compounds in citrus and its relative species as affected by low- and/or high-temperature stress.

‘Cultivar’/Variety/(species)	Tissue	Treatment	Put a	Spd	Spm	Reference
‘Bearss’ Lemon (Citrus limon)	Flavedo	Fruits were stored at chilling temperatures (1°C)	▲	-	-	59
‘Marsh’ Grapefruit (C. paradisi)	Flavedo	Fruits were stored at chilling temperatures (4.4°C)	▲	-	-	59
‘Marsh’ Grapefruit (C. paradisi)	Juice	Fruits were stored at chilling temperatures (4.4°C)	▲	-	-	59
Sour Orange (C. aurantium)	Leaves	Plants were exposed for 3 weeks to cold stress (15.6°C day and 4.4°C night)	▲	▲	▼	9
‘Valencia’ Sweet Orange (C. sinensis)	Leaves	Plants were exposed for 3 weeks to cold stress (15.6°C day and 4.4°C night)	▲	▲	▲	9
Rough Lemon (C. jambhiri)	Leaves	Plants were exposed for 3 weeks to cold stress (15.6°C day and 4.4°C night)	●	▲	▲	9
‘Washington’ Navel Orange (C. sinensis)	Leaves	Plants subjected to 4 weeks of low-temperature (10°C day and 7°C night)	▲	▲	▼	10
‘Marsh’ Grapefruit (C. paradisi)	Flavedo	Fruits were stored at 2°C for 4 weeks ± hot water dips (53°C for 2–3 min)	▲	-	-	60
‘Eureka’ Lemon (C. limon)	Flavedo	Fruits were stored at 2°C for 4 weeks ± hot water dips (53°C for 2–3 min)	▲	-	-	60
‘Sweety’ Oroblanco (C. grandis × C. paradisi)	Flavedo	Fruits were stored at 2°C for 4 weeks ± hot water dips (53°C for 2–3 min)	▲	-	-	60
‘Nagami’ Kumquat (Fortunella margarita)	Flavedo	Fruits were stored at 2°C for 4 weeks ± hot water dips (53°C for 2–3 min)	▲	-	-	60
‘Tahitian’ Lime (C. latifolia)	Fruits	Fruits stored at 0°C	↑↓	-	-	61
‘Emperor’ Mandarin (C. reticulata)	Fruits	Fruits stored at 0°C	↑↓	-	-	61
‘Marsh’ Grapefruit (C. paradisi)	Fruits	Fruits stored at 0°C	↑↓	-	-	61
‘Valencia’ Sweet Orange (C. sinensis)	Fruits	Fruits stored at 0°C	↑↓	-	-	61
`Fortune’ Mandarin (C. reticulata)	Flavedo	Fruits were stored for 45 days at 2°C	▲	▲	●	62
`Fortune’ Mandarin (C. reticulata)	Flavedo	Fruits conditioned for 1, 2, and 3 days at 37°C and 90-95% RH	▲	▲	▲	63
`Fortune’ Mandarin (C. reticulata)	Flavedo	Fruits stored at 2°C for 28 days	▲	↑↓	↑↓	63
`Fortune’ Mandarin (C. reticulata)	Flavedo	Fruits submitted to postharvest hot water dips at 44, 47, 50, 53, 56, and 59°C	▲	▲	▲	64
`Verna’ Lemon (C. limon)	Fruits	Fruits stored at 2°C for 20 days after treatment with putrescine, Ca, or water	▼	▼	↑↓	65
`Fortune’ Mandarin (C. reticulata)	Fruits	Heat-conditioning treatment (HCT) at 37°C with air	▲	▲	▲	66
`Fortune’ Mandarin (C. reticulata)	Fruits	HCT at 37°C, followed by 90 days storage at 2°C	▲	▼	▼	66
`Fortune’ Mandarin (C. reticulata)	Fruits	HCT at 37°C, then 90 days storage at 2°C, then recovering for 7 days at 20°C	▲	▲	▲	66
`Verna’ Lemon (C. limon)	Fruits	Fruits dipped in hot water (45°C) for 10 min, then stored at 15°C for 3 weeks	▲	▲	●	67
Sour orange (C. aurantium)	Leaves	Plants were kept at 6 ± 1°C for 6 days. Samples were taken 5 and 24 hpt.	▲	▲	▼	68
`Jincheng’ Sweet Orange (C. sinensis)	Leaves	Seedlings were placed for 3 hours for high-temperature stress (45°C)	▼	▼	▲	11

^a(▲) signifies for increased levels, (▼) signifies for decreased levels, (●) signifies for no change, (↑↓) signifies for a varied response(s) based on the time of exposure,⁶¹ ripening stage of the fruit,⁶³ and (˗) signifies for not-reported compounds. Abbreviations: Put: putrescine, Spd: spermidine, and Spm: spermine.

Unlike putrescine, spermidine and spermine are less reported in citrus and its relatives and both have been shown a varied response(s) to low- and high-temperatures based on the time of exposure⁶¹ and ripening stage of the fruit.⁶³ For example, spermidine was increased in the leaves of sour orange (C. aurantium), sweet orange (C. sinensis cv. Valencia),andRough Lemon (C. jambhiri) that were exposed for three weeks to low-temperature stress (15.6°C and 4.4°C at daytime and night, respectively)⁹ and Navel Orange (C. sinensis cv. Washington) plants subjected to four weeks of low-temperature stress (10°C and 7 °C at daytime and night, respectively).¹⁰ Likewise, spermidine was increased in fruits (Flavedo) of Fortune’ Mandarin (C. reticulata)^62–64,66 and `Verna’ Lemon (C. limon)<sup>67</sup> after different low- and high-temperature treatments. By contrast, spermidine was decreased in `Verna’ Lemon (C. limon) fruits when stored at 2°C for 20 days after treatment with putrescine, calcium, or distilled water⁶⁵ and `Fortune’ Mandarin (C. reticulata) fruits when stored at 2°C for 90 days after heat-conditioning treatment at 37°C.<sup>66</sup> Similarly, `Jincheng’ Sweet Orange (C. sinensis) leaves had lower levels ofspermidine when seedlings were placed for three hours at 45°C.¹¹

Spermine, regardless of being the minor PA, showed a varied response(s) based on the temperature, plant species,^9,66 time of exposure,⁶¹ ripening stage of the fruit.⁶³ For example, spermine was increased in fruits of `Fortune’ mandarin (C. reticulata) when they conditioned heat-conditioning treatment (HCT) at 37°C with air<sup>63,66</sup> or submitted to postharvest hot water dips between 44–59°C.<sup>64</sup> Spermine accumulated in leaves of `Jincheng’ sweet orange (C. sinensis) seedlings when placed at high-temperature (45°C) for 3 hours¹¹ and in leaves of ‘Valencia’ sweet orange (C. sinensis) and rough lemon (C. jambhiri) when Plants exposure for 3 weeks to cold stress (15.6°C and 4.4°C at daytime and night, respectively).⁹ On the other hand, spermine content was reduced in leaves of sour orange (C. aurantium) and ‘Washington’ navel orange (C. sinensis) when plants exposed to low-temperature stress (10–15.6°C and 4.4–7 °C at daytime and night, respectively) for 3–4 weeks.^9,10 Spermine also decreased in `Fortune’ mandarin (C. reticulata) when fruits conditioned with HCT at 37°C, followed by 90 days of storage at 2°C.<sup>66</sup> Interestingly, the endogenous spermine content fluctuated in the fruits of `Fortune’ mandarin (C. reticulata) and `Verna’ lemon (C. limon) when they stored at 2°C for 20–28 days^63,65

As we mentioned above, because citrus species are native to tropical and subtropical areas, most of the citrus cultivars are susceptible to cold temperatures. Therefore, many previous studies have been carried out to clone and characterize low-temperature stress-responsive genes from citrus and its relative species (Table 2). Although more than 25 genes were cloned successfully from citrus, very few genes have been reported to be involved in PAs-based defensive mechanisms against low-temperature stress.^79,82,88,90 Among the low temperature-responsive genes from citrus, cold-responsive (COR; aka cold-regulated) genes were studied intensively. Most of COR genes are a member of closely related dehydrins (DHNs; late embryogenesis abundant [LEA] protein) from citrus that could potentially contribute to cold tolerance.^{56,69,70,72,74,75,94}

Table 2.

Low-temperature stress-responsive genes that have been cloned from citrus or its relatives.

Genes	GenBank accession No.a	‘Cultivar’/Variety/(species)	References
COR19	L39005	Trifoliate Orange (Poncirus trifoliata)	69
CuCOR19	AB016809	‘Satsuma’ Mandarin (Citrus unshiu)	70
FPAL1	AJ238753	‘Fortune’ Mandarin (C. clementina × C. reticulata)	71
FPAL2	AJ238754	‘Fortune’ Mandarin (C. clementina × C. reticulata)	71
COR15	AY032975	‘Marsh Seedless’ Grapefruit (C. paradisi)	72
CLT (CLTa,CLTb)	AY316308	‘Rubidoux’ Trifoliate Orange (P. trifoliata)	73
CsDHN	AY297793	Navel Orange (C. sinensis)	74
CpDHN	AY160772	‘Star Ruby’ Grapefruit (C. paradisi)	74
CrCOR15	AY327515.1	‘Fortune’ Mandarin (C. clementina × C. reticulata)	75
CuCOR15	AB178479	‘Satsuma’ Mandarin (C. unshiu)	56
PtrPII‐C02	DQ158860	‘Rubidoux’ Trifoliate Orange (P. trifoliata)	76
PI-B05	-	‘Rubidoux’ Trifoliate Orange (P. trifoliata)	77
PI-C10	-	‘Rubidoux’ Trifoliate Orange (P. trifoliata)	77
HOS1	FJ844367	Trifoliate Orange (P. trifoliata)	78
CsSAMDC	FJ496345	‘Newhall’ Navel orange (C. sinensis)	79
CsNAC1	CX306190	‘Cleopatra’ Mandarin (C. reshni)	80
LOS2	GQ144341	Trifoliate Orange (P. trifoliata)	81
PtADC	HQ008237	Trifoliate Orange (P. trifoliata)	82
PtCBF1	DQ790888.1	Trifoliate Orange (P. trifoliata)	83
PtrWRKY2	JX236033.2	‘Rubidoux’ Trifoliate Orange (P. trifoliata)	84
PtrbHLH	JX512645	Trifoliate Orange (P. trifoliata)	85
PtrWRKY1	CX065984	‘Rubidoux’ Trifoliate Orange (P. trifoliata)	86
R2R3MYB	Multiple genes	Sweet orange (C. sinesis	87
R2R3MYB	Multiple genes	Clementine (C. clementina)	87
PtrICE1	KJ812152	Trifoliate Orange (P. trifoliata)	88
CsCBF	FJ861084.1	sweet orange (C. sinensis)	89
Ptr-miR396b	-	Trifoliate Orange (P. trifoliata)	90
p35 anti-apoptotic b	-	From Autographa californica moth Nuclear polyhedrosis virus (AcMNPV) expressed in ‘Carrizo’ citrange (C. sinensis × P. trifoliata)	91,92
CsbHLH18	-	sweet orange (C. sinensis)	93

^a(˗) Signifies for not-reported accession numbers.

For instance, three COR15 genes were cloned and identified including CpCOR15(GenBank accession No. AY032975) from ‘Marsh seedless’ grapefruit (C. paradisi),⁷² CrCOR15(GenBank Accession No. AY327515.1) from ‘Fortune’ mandarin(C. clementina × C. reticulata),⁷⁵ and CuCOR15(GenBank accession No. AB178479) from ‘Satsuma’ mandarin (C. unshiu).⁵⁶ COR15, which is almost completely identical to other reported citrus DHNs proteins, might protect stromal proteins from aggregation under freeze stress conditions via binding metals using a specific sequence containing histidine.⁵⁶ In addition, Two COR19 were cloned from citrus plants including, PtrCOR19 (GenBank accession No. L39005) from trifoliate orange (P. trifoliata)⁶⁹ and CuCOR19 (GenBank accession No. AB016809) from ‘Satsuma’ mandarin (C. unshiu).⁷⁰ Both are DHNs protein and could be orthologues of CsLEA65 with histidine-rich domains.^69,70 Additionally, CuCOR15 and CuCOR19 from C. unshiu contain a relatively considerable amount of histidine, arginine, and other reactive amino acid residues on the surface of their folded molecules and exhibited a ROS-scavenging activity and metal ion binding properties.^56,70 Moreover, two DHNs were cloned including CsDHN from navel orange (C. sinensis)and CpDHN from ‘star ruby’ grapefruit (C. paradisi) (GenBank accession No. AY297793 and AY160772, respectively).⁷⁴ Interestingly, CpDHN mRNA levels were up-regulated in the fruit peel tissue when fruits exposed to short hot‐water treatment followed by storage at chilling temperatures.⁷⁴ These findings suggest that DHNs might play a molecular regulatory role in heat‐induced chilling tolerance in citrus plants.

Unfortunately, the interaction between COR/DHNs genes and PAs in citrus is poorly studied and, to our knowledge, it is not reported yet. However, PAs were reported to be associated with DHN genes to regulate tolerance to water stress in other plant species such as white clover (Trifolium repens). Exogenously applied spermidine (20 μM) to detached leaves of T. repens significantly increased the transcript levels of three DHNs genes including, Dehydrin (Y₂SK), DHN1 (Y₂ K), and Dhn b (SK₂) at different times.⁹⁵ Nevertheless, further studies are required to explore the relationship between COR/DHNs genes and PAs in citrus plants.

In addition to CORs/DHNs genes, two full-length cDNA clones of _L-phenylalanine ammonia-lyase (PAL) were isolated from Fortune mandarin (C. clementina × C. reticulata) and described as FPAL1 and FPAL2 (GenBank accession No. AJ238753 and AJ238754, respectively).⁷¹ PAL is generally recognized as a marker of environmental stress in different plant tissues and could be involved in PAs-based defensive response of citrus fruits and other organs to cold temperatures. Interestingly exogenous PAs application significantly enhanced the PAL activity in Scots pines (Pinns sylvestris L.) callus in vitro.⁹⁶ On the other hand, inhibition of PAL activity in alfalfa (Medicago sativa) cell suspension culture let to a remarkable reduction in endogenous PAs content.⁹⁷ Taken together, these findings demonstrated the potential role of FPAL1 and FPAL2 in citrus responses to low-temperature stress.

Furthermore, several low temperature-responsive genes were cloned and identified from trifoliate orange (P. trifoliata), an extremely cold hardy citrus rootstock. These genes including citrus low-temperature (PtrCLT; GenBank accession No. AY316308) gene with two transcripts, called CLTa and CLTb,⁷³ PtrPII‐C02 (a really interesting gene- [RING‐] Zn finger gene; GenBank accession No. DQ158860),⁷⁶ PtrPI-B05 (containing an apetala2/ethylene response factor (AP2/ERF) domain and showed homology with ERF proteins from other plants)and PtrPI-C10 (containing both AP2/ERF and B3 DNA binding domains),⁷⁷ high expression of osmotically responsive genes 1 (PtrHOS1;GenBank Accession No. FJ844367),⁷⁸ Low expression of osmotically responsive genes 2 (PtrLOS2; GenBank accession No. GQ144341),⁸¹ C-repeat/dehydration-responsive elementbinding factor (PtCBF1; GenBank accession No. DQ790888.1),⁸³ and two WRKY transcription factors described as PtrWRKY1 and PtrWRKY2(GenBank accession No. CX065984 and JX236033.2, respectively)^84,86, Interestingly, CBFs bind CRT/DRE motifs that exist in the promoters of numerous COR genes to activate gene transcription⁹⁸ and might interact with Inducer of CBF Expression 1 (ICE1) gene.⁸⁸

Likewise, several low temperature-responsive genes were cloned and identified from other citrus species including CsNAC1transcription factor (GenBank accession No. CX306190) from C. reshni,⁸⁰ multiple R2R3MYBgenes from C. sinesis and C. clementina,⁸⁷ CsCBF (GenBank accession No. FJ861084.1) from C. sinensis,⁸⁹ and the basic helix–loop–helix (CsbHLH18) transcription factor from C. sinensis.⁹³ Some of these genes might play a key regulatory role in PAs-based defensive response(s) of citrus plants against chilling, however, more studies are required to clarify this role.

Wang et al. (2010 and 2011) cloned two genes involved in PAs biosynthesis includingCsSAMDC (GenBank Accession No. FJ496345) from navel orange (C. sinensis)⁷⁹ andPtADC (GenBank accession No. HQ008237) from trifoliate orange(P. trifoliata).⁸² CsSAMDC is a rate-limiting enzyme for PAs biosynthesis and the temporal alteration in its transcript levels may be part of the PAs-based defensive response(s) of citrus plants since its mRNA levels varied in different citrus genotypes with variable levels of tolerance to salinity or to low-temperature stress.⁷⁹ In addition, SAMDC (aka AMD1) is a key enzyme that connects both ethylene biosynthesis and PAs biosynthesis pathways. Ethylene plays diverse roles in abiotic stress tolerance,⁹⁹ however, the molecular mechanisms behind ethylene-SAMDC-PAs defensive role(s) remain poorly understood. Likewise, PtADC is a key enzyme responsible for PAs biosynthesis under stress conditions, which might be involved in the citrus response to low temperature, since its transcript levels were up-regulated by low temperature.⁸² Moreover, the PtADC overexpressing mutant of A. thaliana (adc1-1) had a higher PAs content, particularly putrescine, and showed enhanced resistance to several abiotic stress types such as dehydration, drought, and cold stress.⁸² In addition, the adc1-1 transgenic line accumulated less ROS under different stresses but exhibited higher ROS scavenging capacity.⁸² Interestingly, the ROS scavenging capacity ofthe adc1-1transgenic line was reduced significantly when plants were treated with _D-arginine (an ADC inhibitor) prior to stress treatment.⁸² Taken together, these findings prove that both genes (SAMDC and ADC) are involved in PAs-based defensive responses against low-temperature stress, however, their exact role remains ambiguous and required more studies.

Furthermore, two ICE homologs have been isolated and cloned from trifoliate orange (P. trifoliata) plants including, PtrbHLH (GenBank accession No. JX512645)⁸⁵ and PtrICE1 (GenBank accession No. KJ812152).⁸⁸ Both homologs were up-regulated upon cold stress. Overexpression of PtrbHLH or PtrICE1 in tobacco (N. tabacum) or lemon (C. limon) resulted in enhanced tolerance to cold stresses at either chilling (around 4°C) or freezing (0 °C) temperatures^85,88 but through two different mechanisms. PtrbHLH is a cold-responsive transcription factor that enhances cold tolerance in citrus plants via activating the POD-mediated and ROS scavenging of hydrogen peroxide (H₂O₂), ⁸⁵ while PtrICE1 encodes an MYC-like bHLHtranscription factor that functions as a central regulator of PAs-based defensive response against low-temperature via modulating endogenous PAs levels through interacting with PtrADC.⁸⁸ Yeast two-hybrid screening revealed that PtrADC was confirmed as a bona fide protein interacting with PtrICE1. Furthermore, overexpression of PtrICE1 in tobacco (N. tabacum) and lemon (C. limon) plants exhibited slightly higher transcript levels of ADC under normal growth condition but significantly induced upon cold stress, consistent with changes in endogenous PAs content, which resulted in enhanced tolerance to cold stresses at either chilling (4°C) or freezing (0°C) temperatures.⁸⁸

Taken together, these findings suggest that ICE1–CBF–COR signaling cascade plays a critical role in mediating PAs-based defensive responses against low-temperature stress. ICEs functions as a master regulator of this cascade via controlling the expression of CBFs, which sequentially regulate the downstream cold-responsive genes.^100,101 For example, in A. thaliana, ICE homologs (AtICE2, and AtICE1) were reported to be involved in cold tolerance via modulation of CBF1 transcript levels.¹⁰² Later, and during cold acclimation, CBFs activate COR genes and resulting in the acquisition of low-temperature tolerance.^98,103 Furthermore, ICE1 acts as a central regulator of PAs-based defensive response against low-temperature via modulating endogenous PAs content through interacting with PtrADC,⁸⁸ which adds a new avenue for the ICE1–CBF–COR signaling cascade. However, the complete role of ICE1 in citrus plants is not fully understood, and many other ICE1-based mechanisms remain unexplored. For instance, the ICE1-mediated cold response appeared to be positively regulated by the phytohormone jasmonate (JAs) in Arabidopsis plants via regulation of the ICE-CBF/DREB1 transcriptional cascade,¹⁰⁴ though, this role is not studied in citrus yet. Recently, Geng and Liu (2018) showed that the transcription factor CsbHLH18is involved in cold tolerance and ROS homeostasis via the regulation of antioxidant enzymes of C. sinensis.⁹³ Interestingly, CsbHLH18is a potential ICE homolog, thatis localized in the nuclei and has transcriptional activation activity.⁹³ Furthermore, CsbHLH18 directly and specifically binds to the promoter of CsPOD and activates it.⁸⁵

In addition to all cloned genes, several cold-responsive microRNAs (miRNA) have been computationally identified, however, the functions of most of them are poorly explored. Zhang et al. (2016) investigated the function of miRNA396b, a cold-responsive miRNA, from trifoliate orange (Ptr-miR396b) in citrus response to cold stress and its potential regulatory role in ethylene–PAs homeostasis.⁹⁰ Overexpression of Ptr-MIR396b, the precursor of Ptr-miR396b, in lemon (C. limon) plants showed higher endogenous free PAs content (putrescine, spermidine, and spermine) and their biosynthetic genes (SAMDC, SPDS, and SPMS), but lower transcript levels of ACO and ethylene content than wildtype plants under cold stress.⁹⁰ These findings suggest that Ptr-miR396bplay a key role in citrus response to cold stress via regulating ACO and modulating ethylene-PAs homeostasis⁹⁰

In addition, previous studies provide strong evidence that PAs confers tolerance to low-temperature stressvia modulation of the antioxidative capacity of citrus plants.^{56,58,88,90,105} Early this century, Sala and Lafuente (2000) showed that catalase (CAT) enzyme activity was associated with greater chilling tolerance in ‘Fortune’ mandarins when the fruits conditioned for 3 days at 37°C with air, then stored at 2°C. Moreover, exogenous treatment with 3-amino-1,2,4-triazole (AT; a chemical inhibitor of CAT activity) reduced the enzymatic activity of CAT in the AT-treated fruits, but not ascorbate peroxidase (APX), glutathione reductase (GR) and superoxide dismutase (SOD). The reduction of CAT activity was associated with greater peel damage in ‘Fortune’ mandarins and in the chilling-tolerant ‘Clementine’ and ‘Clemenules’ cultivars stored at 2°C but not at 12°C.¹⁰⁵ Recently, it was reported that the foliar application of PAs enhanced the CAT activity in Bakraii citrus.³⁵ Likewise, exogenous spermine treatment significantly increased the activities of CAT, SOD, and peroxidase enzymes, but reduced both H₂O₂ and superoxide (O₂⁻) in spermine-pretreated trifoliate orange seedlings when exposed to both drought and high-temperature (45°C).⁵⁸

Furthermore, the adc1-1 mutant of A. thaliana (a PtADC-overexpressing mutant) exhibited higher PAs content, accompanied by higher ROS scavenging capacity and reduced ROS accumulation in cold-stressed plants, which was reversed when the transgenic lines were treated with the ADC inhibitors.⁸² Additionally, the overexpression of PtrICE1 that encodes a MYC-like bHLH transcription factor in tobacco (N. tabacum) or lemon (C. limon) showed higher PAs content, up-regulated PtADC transcript levels, and higher activities of antioxidant enzymes (SOD and CAT), but appreciably alleviated the accumulation of ROS (H₂O₂ and O₂^–) in the transgenic lines under cold conditions.⁸⁸ Likewise, overexpression of CsbHLH18 in the same model plants resulted in peroxidase (POD) activity, but significantly less ROS in the transgenic lines after cold treatment.⁸⁵ On the other hand, knockdown of bHLH18 by RNAi in P. trifoliata conferred lower POD activity, but higher H₂O₂ accumulation accompanied by enhanced cold susceptibility while exogenous POD supplementation improved cold tolerance of the RNAi line which was inversely correlated with the H₂O₂ levels.⁸⁵ Interestingly, CsbHLH18 directly and specifically binds to the promoter of CsPOD and activates it.⁸⁵ Similarly, overexpression of Ptr-MIR396b, the precursor of Ptr-miR396b, in C. limon plants showed reduced ROS accumulation consistent with higher PAs content and up-regulation of their biosynthetic genes and positively regulates cold tolerance via reducing the expression of ACO and suppressing ethylene biosynthesis.⁹⁰

A previous model has been proposed for the integration of PAs biosynthesis with ABA-dependent and ABA-independent signaling pathways in the cold acclimation response.⁹⁸ According to Alcázar’s model, cold stress increases the transcript levels of ADC1, causing significant induction in PAs biosynthesis, particularly putrescine. The high endogenous putrescine content might initiate the ABA-dependent signaling pathway via inducing 9-cis-epoxycarotenoid dioxygenase (NCED3) resulting in ABA accumulation, which, subsequently, activate the abscisic acid-responsive element (ABRE)-COR. Later, the activation of the ABA-dependent signaling pathway resulting in a cold acclimation response after cold treatment.⁹⁸ Interestingly, exogenous ABA treatment induces ADC2 transcription in an ABA-dependent manner.¹⁴ Additionally, cold stress could activate the ABA-independent signaling pathway via inducing the C-repeat (CRT)/dehydration-responsive element (DRE)-binding factors (CBF/DREB1) proteins using ICE1. CBFs bind CRT/DRE motifs in the promoters of COR genes to activate gene transcription and play a key role in cold acclimation.⁹⁸ The induction of ABA-responsive element (ABRE) binding factor and NCED3 (a major ABA-biosynthetic gene) in spermine-treated trifoliate orange seedlings⁵⁸ support the idea that PAs are integrated with ABA to establish citrus tolerance against low- and high-temperature. However, the PAs-based defensive mechanisms against low-temperature stress are more complicated than PAs-ABA interaction.

Therefore, we proposed a new hypothetical model for the PAs-based defensive mechanisms against low-temperature (Figure 1). In addition to the integration of PAs biosynthesis with ABA-dependent and ABA-independent signaling pathways suggested byAlcázar et al. 2011,⁹⁸ we suggest that citrus plants develop a multi-layered defensive system to protect themselves. The defense mechanisms include PAs, phytohormones (such as ethylene, ABA, and JAs), accumulation of low-temperature-responsive genes, alteration of the plant signaling system, induction of some miRNA, and modulation of ROS scavenging capacity by regulating the activities of antioxidant enzymes.

Figure 1.

Hypothetical model of the potential polyamines- (PAs-) based responses of citrus plants to cold stress. (a) Integration of PAs with other metabolic pathways in cold acclimation of citrus plants, and (b) Activation of C-repeat/dehydration-responsive element binding factor (CBF) and cold-responsive (COR; aka cold-regulated) genes. Solid lines with arrowhead signify positive reaction, dashed lines with blunt-heads indicate negative interaction, while round-dotted lines with arrows represent hypothetical mechanisms or uncharacterized elements. For more details and abbreviations, see the main text.

According to our model, cold stress induces the PAs’ biosynthetic genes (particularly ADC), causing a significant increase in endogenous PAs content, particularly putrescine. The high endogenous putrescine level directly regulating the antioxidant enzymes,⁹³ leading to enhanced ROS scavenging (Figure 1(a)). Additionally, endogenous PAs levels could be modulated by ICE1 via interacting with ADC.⁸⁸ Interestingly, ICEs play a key regulatory role in the activation of CBFs (Figure 1(b)) via binding to the MYC-recognition motif in the CBF promoter.^100,101,106 Sequentially, CBFs activate the transcription of COR genes through binding to CRT/DRE motifs that exist in the gene promoter(Figure 1(b)).⁹⁸ Interestingly, the expressions of ICEs, particularly ICE1, are negatively regulated by HOS1-E3 ubiquitin ligase¹⁰⁶ or mitogen-activated protein kinase (MKK4/5-MPK3/6) cascade.¹⁰⁷ HOS1 ubiquitinates the ICE1, while MPK3 and MPK6 can phosphorylate ICE1, which sequentially promotes the degradation of ICE1.^106,107

Furthermore, the high endogenous PAs content might initiate both ABA-dependent and ABA-independent signaling pathway via inducing NCED3 and ICE1, respectively (as described by Alcázar et al., 2011). ABA-dependent signaling pathway enhances the accumulation of ABA and activation of ABRE-CBF, whilethe ABA-independent signaling pathway induces the activation of CBF/DREB1.⁹⁸ Both pathways resulting in a cold acclimation response via regulating the transcription of COR genes. In addition to ABA pathways, other phytohormones (such as ethylene and JAs) might be involved in PAs-based defensive mechanisms against low-temperature. Briefly, JAs regulate the ICE-CBF/DREB1 transcriptional cascade¹⁰⁴ via binding to F-box protein coronatine-insensitive 1 (COI1), while the ethylene pathway supports the PAs biosynthesis via SAMDC (aka AMD1).⁷⁹ SAMDC is a stress-responsive gene that its transcript levels varied in different citrus genotypes with variable levels of tolerance to salinity or to low-temperature stress⁷⁹ and highly expressed at the initial stages of either stress treatment in tolerant genotypes.

Moreover, miRNA (such as Ptr-miR396b) might be involved in PAs-based defensive mechanisms against low-temperature. Ptr-miR396b plays a regulatory role in ethylene–PAs homeostasis via regulating the expression of ACO gene.⁹⁰ Ptr-miR396b negatively regulates the transcript levels of ACO, causing lower ethylene content, but positively regulates the PAs-biosynthetic genes (SAMDC, SPDS, and SPMS), resulting in higher endogenous free PAs content (putrescine, spermidine, and spermine).⁹⁰

Several previous studies should that role of PAs against low-temperature involve the activation ofCa²⁺influx^108–111 and induction of PAL enzymes.^71,96,97 For example, cytosolic Ca²⁺ levels were increased within seconds of a cold shock in tobacco,¹⁰⁸ which was associated with the transcript levels of COR gene.^110,111 Plants possess several sensors for the cold-induced Ca²⁺signature, including calcium-modulated protein (CaM, aka calmodulin) and Ca²⁺–dependent protein kinase (CDPK).¹¹¹ Calmodulin is a negative regulator of COR genes,¹¹² while CDPK is a positive regulator of COR genes and cold tolerance in plants.¹⁰⁹ Furthermore, PAL could be involved in the citrus response to cold stress⁷¹ via modulating lignin biosynthesis¹¹³ and/or endogenous PAs contents. The positive correlation between PAL and PAs content was reported in other plant species such as P. sylvestris⁹⁶ and M. sativa.⁹⁷

Taken together, citrus plants develop a PAs-based multi-layered defensive system to protect themselves against low-temperature stress. This defensive system requires the orchestration of biochemical, metabolic, transcriptional, and physiological modifications to establish the cold acclimation. Mainly, this defensive system relay on the modulation of ROS homeostasis via the enhancement of antioxidant defense directly and/or the induction of the low temperature-responsive gene expressions. Many genes and transcription factors are implicated in the regulation of ROS homeostasis such as CORs,^56,69,70 CsNAC1,⁸⁰ PtrbHLH,⁸⁵ PtrICE1,⁸⁸ CsbHLH18,⁹³ and the miRNA Ptr-miR396b.⁹⁰ However, the molecular mechanism(s) underlying the PAs-based defensive system against low-temperature is still limited and more investigations are required to gain a better understanding of its regulatory mechanisms.

Role of PAs in citrus response(s) to salinity

More than one-third of irrigated soils worldwide are affected by salt stress (aka salinity).^114,115 Salinity is one of the most important environmental factors that threating the growth of many crop plants and reducing their productivity.^116,117 Citrus and most of their relative species have been reported previously as salt-sensitive plants.^118–122 However, citrus rootstocks are varied in their salinity tolerance,¹¹⁵ and their relative tolerance could be affected by climate conditions, fertilization and nutrition, soil type, irrigation methods, and the genetic background of the rootstock itself.^115,122 In citrus, salinity causes various physiological and biochemical alterations due to the concentrations of chloride anion (Cl⁻) and sodium cation (Na⁺) in leaves and other plant organs.¹²¹ In addition, soil salinization reduces citrus growth, decreases the fruit yield and quality,¹²³ causes nutrient deficiencies or imbalances,¹¹⁵ and impairs photosynthetic activity.¹²⁴ In addition, salt stress led to a significant reduction in the gas-exchange and chlorophyll fluorescencecharacteristicsin Bakraii (C. reticulata × C. limetta) seedlings including relative chlorophyll contents, chlorophyll fluorescence yields (F_v/F_m), net photosynthetic and respiration rates,⁵¹ net photosynthetic rate (P_N), stomatal conductance (g_s), and intercellular CO₂ concentration (C_i).³⁵ Nevertheless, exogenous application of different PAs, regardlessthe concentrations, reversed the negative effects of salinity and enhanced the overall growth of citrus plants.^{34,35,47,124–126} Moreover, the PAs-treated citrus plants maintained higher gas-exchange and chlorophyll fluorescencecharacteristics when exposed to salt stress.³⁵

Generally, most of the previous studies showed that putrescine and spermidine fluctuated and showed a varied response to salt stress (Table 3).^{34,45–47,49-52} The varied effects of salinity on the putrescine and spermidine content could be due to the variation in soil type, application method, and the rootstocks. For example, in non-stressful conditions, the salt-tolerant rootstock ‘Cleopatra’ mandarin (C. reshni) had high spermidine and spermine but lower putrescine compared with the salt-sensitive cultivar ‘Troyer’ citrange (P. trifoliata × C. sinensis).⁴⁹ However, under salinity conditions, putrescine was increased and spermidine and spermine were decreased in salinity-stressed plants.^49,50 On the other hand, in most cases, spermine was induced to higher levels after salt stress (Table 3), ^{34,45–47,51,52} which indicates that the PAs-based defensive mechanisms against salinity stress in citrus mainly depends on spermine, but not other PAs compounds. The PAs induction has been suggested to be a metabolic response to cytoplasm acidification due to anion/cation imbalance caused by salt stress.⁵² Another point of view for the PAs induction was supported by the ROS-scavenging activities of PAs which can help the stressed plants to overcome the stress-activated physiological damage.⁵² In addition, the phytohormones ABA, ethylene, and ethylene precursor (ACC; 1-Aminocyclopropane-1-carboxylic acid) were increased after salt stress.^46,49,50,127 For instance, the salt-sensitive cultivar ‘Troyer’ citrange (P. trifoliata × C. sinensis) produced ethylene and its precursor (ACC) at a higher rate than the chloride-tolerant rootstock `Cleopatra’ mandarin (C. reshni).⁵⁰ Taken together, these findings indicate that PAs-based defensive mechanisms against salt stress could be associated with the phytohormones-based defensive system, however, the molecular mechanisms behind this correlation are poorly studied in citrus plants.

Table 3.

Changes in the endogenous content of major PAs, proline (Pro), abscisic acid (ABA), ethylene (ET), and 1-aminocyclopropane-1-carboxylic acid (ACC; ethylene precursor) in citrus and its relative species as affected by salinity/chloride stress.

‘Cultivar’/Variety/(species)	Tissue	Treatment	Put a	Spd	Spm	ET	ACC	ABA	Pro	Reference
‘Cleopatra’ mandarin (Citrus reshni)	leaves	The plants were irrigated with nutrient solutions containing 2, 16, or 48 mM chloride (w/o nitrate) fora 3-month experimental period.	▲	●	▼	▲	▲	˗	˗	49,50
‘Troyer’ citrange (P. trifoliata × C. sinensis)	leaves	The plants were irrigated with nutrient solutions containing 2, 16, or 48 mM chloride (w/o nitrate) for 3-month experimental period.	▲	▼	▼	▲	▲	˗	˗	49,50
‘Carrizo’ citrange (C. sinensis × P. trifoliata)	leaves	Seedlings were not progressively adapted to saline conditions but treated with salt shock by adding 200 mM NaCl to a modified Hoagland solution.	-	-	-	▲	▲	▲	▲	127
	Roots		-	-	-	˗	▲	▲	˗	127
‘Cleopatra’ mandarin (Citrus reshni)	Roots	Seedlings were irrigated with half-strength Hoagland nutrient solution containing 0, 40 or 80 mM NaCl for 12 weeks.	▼	▼	▼	˗	˗	˗	▲	51
	leaves		▼	▼	▲	˗	˗	˗	↑↓	51
‘Troyer’ citrange (P. trifoliata × C. sinensis)	Roots Leaves	Seedlings were irrigated with half-strength Hoagland nutrient solution containing 0, 40 or 80 mM NaCl for 12 weeks.	▼	▼	●	˗	˗	˗	↑↓	51
			▼	↑↓	▲	˗	˗	˗	▲	51
‘Newhall’ navel orange (C. sinensis)	embryogenic callus	The 21-day old callus was transferred to callus growth medium with 100 mM NaCl and maintained at 25ºC for 7 days.	▲	▲	▲	˗	˗	˗	˗	52
‘Cleopatra’ mandarin (Citrus reshni)	leaves	Seedlings were grown in Hoagland solution containing 75 mM NaCl	▲	●	▲	˗	˗	˗	▲	45
‘Cleopatra’ mandarin (Citrus reshni)	leaves	The same previous saline solution + 1 mM D-arginine (PAs inhibitor)	●	●	▲				▲	45
‘Troyer’ citrange (P. trifoliata × C. sinensis)	leaves	Seedlings were grown in Hoagland solution containing 75 mM NaCl	▲	●	▲	˗	˗	˗	▲	34
‘Troyer’ citrange (P. trifoliata × C. sinensis)	leaves	The same previous saline solution + 0.1 mM spermidine	●		▲				▲	34
Ten citrus rootstocks b	leaves	Seedlings were acclimated for 90 days, the four levels of NaCl (0, 30, 60 and 90 mM) were applied along with half-strength Hoagland’s solution for another 90 days.	˗	˗	˗	˗	˗	˗	▲	128
Karna khatta sour orange hybrid (C. karna)	leaves	Seedlings were grown in soil with EC around 3.0 dS m−1 vs. garden soil about 0.35 dS m−1 (as control)	˗	˗	˗	˗	˗	˗	▲	125
‘Carrizo’ citrange (C. sinensis × P. trifoliata)	leaves	Seedlings were irrigated with Hoagland solution + 90 mM NaCl for 2 months.	▼	▼	●	˗	˗	▲	▲	46
‘Carrizo’ citrange (C. sinensis × P. trifoliata)	leaves	The same previous saline solution, w/o 5 mM of N-NH4+ (NH4+ treatment)	▲	▲	▲	˗	˗	▲	▲	46
Sour orange (C. aurantium)	leaves	Seedlings were treated with NaCl via soil drench (0.0 vs. 150 mM)	˗	˗	˗	˗	˗	˗	▲	126
Sour orange (C. aurantium)	leaves	Seedlings were exposed to 100 mM NaCl + 1 mM exogenous PAs (putrescine, spermidine, or spermine) added to half-strength Hoagland’s nutrient solution	▼	●	▲	˗	˗	˗	▲	47
Sour orange (C. aurantium)	Roots		●	▼	●	˗	˗	˗	˗	47
nine citrus rootstocks c	leaves	Seedlings were acclimated for 5 months, then treated with four levels of NaCl (0, 25, 50 and 75 mM). NaCl concentration was increased gradually to avoid osmotic shock	˗	˗	˗	˗	˗	˗	▲	129
‘Bakraii’ mandarin (C. reticulata × C. limetta)	leaves	Seedlings were acclimated for 6 months, then treated with 75 mM NaCl at 3-days intervals using 0.5 L of irrigation water. NaCl concentration was increased gradually to avoid osmotic shock.	˗	˗	˗	˗	˗	˗	▲	35

^a(▲) signifies for increased levels, (▼) signifies for decreased levels, (●) signifies for no change, (↑↓) signifies for varied responses, and (˗) signifies for not-reported compounds.Abbreviations: Put: putrescine, Spd: spermidine, Spm: spermine, ABA: abscisic acid, ET: ethylene, ACC: 1-aminocyclopropane-1-carboxylic acid, and Pro:_L-proline.

^bThey tested ten citrus rootstocks included: Rough lemon (C. jambhiri), Rubidoux trifoliate (P. trifoliata), Citrus Obvidea, Rangpur Lime (C. limonia), Sanchton Citrumillo (X Citroncirus), Citrumillo-1452 (X Citroncirus), Carrizo Citrange (C. sinensis × P. trifoliata), Yuma Citrange (P. trifoliata × C. sinensis), Gada Dehi, and Bitter Sweet Orange (C. aurantium).¹²⁸

^cThey tested nine citrus rootstocks included: Sour orange (C. aurantium), Bakraii (C. reticulata ×C. limetta), Cleopatra mandarin (C. reshni), Rangpur lime (C. limonia), Rough lemon (C. jambhiri), Macrophylla (C. macrophylla), Swingle citrumelo (P. trifoliata ×C. paradisi), Citrange (P. trifoliata ×C. sinensis) and Trifoliate orange (P. trifoliata).¹²⁹

Furthermore, the amino acid _L-proline usually induced to higher levels in citrus leaves after salt stress (Table 3).^{34,35,45–47,51,125–129} Proline is a common osmolyte multifunctional amino acid¹³⁰ and also a major signaling molecule and plant growth regulator that triggers cascade signaling processes.¹³¹ In various plant species, proline was reported to be involved in crop tolerance against various abiotic stresses particularly salinity by maintaining ROS homeostasis via modulation of antioxidant enzymes.^130–133 Nevertheless, the precise functional mechanisms of proline as an osmolyte in citrus plants are still ambiguous. Proline may function as ROS scavenger, osmotic adjuster, guardian for enzymes and DNA from salt damage(s) and/or signaling/regulatory molecule.¹³⁴ For example, exogenous application of proline (5 mM) to cultured salt-sensitive Citrus sinensis ‘Valencia late’ cell line positively influenced the symptoms of salinity stress (100 mM NaCl) and increased growth of this cell line.¹³⁴ Likewise, the addition of proline (particularly at a concentration of 25 mg L⁻¹) to the basal culture media significantly increased the plantlet height and leaf number of C. sinensisplantlets and alleviated the inhibitory effect of salt stress.¹³⁵ Interestingly, Aziz et al. (1998) reported that stress‐induced changes in PAs contents could regulate the proline accumulation in tomato leaf discs under salt stress.¹³⁶

Moreover, overexpression of the salt-induced small protein (FcSISP; aka F. crassifoliaStress-Inducible Small Protein) from Meiwa kumquat (Fortunella crassifolia) in trifoliate orange (P. trifoliata) and tobacco (N. nudicaulis)conferred enhanced tolerance to salinity stress in the transgenic plants.¹³⁷ This higher tolerance to salinity was accompanied by higher levels of proline in tobacco transgenic plants.¹³⁷ Additionally, transcript levels of two major proline biosynthetic genes including delta 1-pyrroline-5-carboxylate synthetase (P5CS) and pyrroline-5-carboxylate reductase (P5CR) were significantly higher in the N. nudicaulistransgenic plants.¹³⁷ Interestingly, FcSISP was quickly induced after the salt stress (200 mM NaCl) and rose progressively until reaching its highest peak at 24 hour-post-treatment (hpt), but it mildly induced by other stressors such as dehydration, cold, and even upon the infection with the bacterial phytopathogen causing citrus canker (Xanthomonas citri subsp. citri [Xcc]).¹³⁷ Together, these findings indicate that FcSISP may acts as a regulator for the proline biosynthesis in citrus upon salinity stress.

In comparison with other fruit trees, citrus species are considered the most salt-sensitive plants^118-122 which greatly limit the citrus growth worldwide. Therefore, many previous studies have been carried out to clone and characterize salinity-responsive genes from citrus and its relative species (Table 4). Although about 22 genes were cloned successfully from citrus and two genes were genetically transformed to citrus,^142–144 however, very few genes have been reported to be involved in PAs-based defensive mechanisms against salt stress. For example, ‘Carrizo’ citrange (C. sinensis × P. trifoliata) and Rough lemon (C. jambhiri), are the most commonly used citrus rootstocks worldwide, however, both are very sensitive to salinity stress, which restricts their use in salty soils.^142,144 Overexpression of the halotolerance (HAL2)gene fromSaccharomyces cerevisiae in ‘Carrizo’ citrange¹⁴² and Rough lemon¹⁴⁴ improved the tolerance to salinity in transgenic plants.^142,144 HAL 2 gene is in the methionine biosynthetic pathway, that encodes for a salt-sensitive biphosphate nucleotidase, which is required for sulfate accumulation in S. cerevisiae.¹⁵⁵ HAL 2 gene is involved in salt tolerance mechanism, where it confers tolerance to lithium and sodium ions. Likewise, overexpression of a betaine aldehyde dehydrogenase (AhBADH) gene from Atriplex hortensis in trifoliate orange (P. trifoliata) enhanced the tolerance to salinity in transgenic lines, which may be correlated with the low levels of lipid peroxidation, protection of the photosynthetic machinery, and increase in K⁺ uptake.¹⁴³

Table 4.

Salinity-responsive genes that have been cloned/genetically transformed from/to citrus or its relatives.

Genes	GenBank accession No. a	‘Cultivar’/Variety/(species)	References
Cit-SAP	-	‘Shamuti’ Sweet orange (C. sinesis)	138
CSA (Cit-SAP)	-	‘Shamuti’ Sweet orange (C. sinesis)	139
CsLea5	-	Sweet orange (C. sinensis)	140
C3 (Oleosin 1)	Z48450.1	Sweet orange (C. sinensis)	141
HAL2 b	X72847	From yeast (Saccharomyces cerevisiae) expressed in ‘Carrizo’ citrange (C. sinensis × P. trifoliata)	142
CsSAMDC	FJ496345	‘Newhall’ Navel orange (C. sinensis)	79
CsNAC1	CX306190	‘Cleopatra’ Mandarin (C. reshni)	80
AhBADH b	-	From Atriplex hortensis expressed in P. trifoliata	143
HAL2 b	X72847	From yeast (Saccharomyces cerevisiae) expressed in Rough lemon (C. jambhiri)	144
PIP1 & PIP2	-	‘Cleopatra’ mandarin (C. reshni) ‘Carrizo’ citrange (C. sinensis × P. trifoliata) Trifoliate Orange (P. trifoliata)	145
CsCLCc	GU942490.1	Trifoliate Orange (P. trifoliata)	146
FcSISP	KJ819954.1	Meiwa kumquat (Fortunella crassifolia)	137
CrNCED1	DQ028471.1	‘Cleopatra’ mandarin (C. reshni)	147
AtCBF3/DREB1A b	At4g25480	From Arabidopsis thalianaexpressed in C. macrophylla	148
CisAL7	Cs3g01400	Sweet orange (C. sinensis)	149
PtrICE1	KJ812152	Trifoliate Orange (P. trifoliata)	88
CsPAO1	Cs7g02060.1	Sweet orange (C. sinensis)	150
CsPAO2	Cs7g18840.2	Sweet orange (C. sinensis)	150
CsPAO3	Cs6g15870.1	Sweet orange (C. sinensis)	150
CsPAO4	Cs4g141 50.1	Sweet orange (C. sinensis)	150,151
CsPAO5	Cs7g23790.1	Sweet orange (C. sinensis)	150
CsPAO6	Cs7g23670.1	Sweet orange (C. sinensis)	150
Four CsWRKYs(Cs30, Cs40, Cs41 and Cs48)	-	Sweet orange (C. sinensis)	152
Four CcWRKYs(Cc53, Cc54, Cc60, and Cc67)	-	Clementine (C. clementina)	152
CsCBF	FJ861084.1	Sweet orange (C. sinensis)	89
PtrZPT2-1	KC820894	Trifoliate orange (P. trifoliata)	153
p35 anti-apoptotic b	-	From Autographa californica moth Nuclear Polyhedrosis Virus (AcMNPV) expressed in ‘Carrizo’ (C. sinensis × P. trifoliata)	91,92
PtrCCC	GU942489.1	Trifoliate orange (P. trifoliata)	154

^a(˗) Signifies for not-reported accession numbers.

^bThese genes were genetically transformed into citrus species using Agrobacterium tumefaciens-mediated gene transfer.

Moreover, several genes were reported to be involved in citrus response(s) to salt stress (Table 4). These genes including salt-associated protein (Cit-SAP) ¹³⁸ andCSA (a gene encoding for Cit-SAP)¹³⁹ from ‘Shamuti’ Sweet orange (C. sinesis) and the plasmamembrane intrinsic proteins (PIP1 and PIP2) from ‘Cleopatra’ mandarin (C. reshni), ‘Carrizo’ citrange (C. sinensis × P. trifoliata) and Trifoliate Orange (P. trifoliata).¹⁴⁵ Likewise, three salinity-responsive genes were cloned successfully from Trifoliate Orange (P. trifoliata) including a putative Chloride channel (CLC)-encoding gene (CsCLCc), ¹⁴⁶ Cys2/His2 (C2H2) zinc finger proteins (PtrZPT2-1),¹⁵³ and Cation-chloride cotransporters membrane protein (PtrCCC).¹⁵⁴ Furthermore, late-embryogenesis 5 (C-Lea5) gene,¹⁴⁰ Oleosin 1 (aka C3),¹⁴¹ Alfin1-like (CisAL7),¹⁴⁹ sixpolyamine oxidase (CsPAO)genes(CsPAO1 – CsPAO6), ^150,151 and four CsWRKYs(Cs30, Cs40, Cs41 and Cs48)¹⁵² were cloned from sweet orange (C. sinensis) and identified as salinity-responsive genes.

Moreover, Gong et al. (2014) isolated a stress-responsive gene from Meiwa kumquat (F. crassifolia)(described as FcSISP).¹³⁷ Overexpression of FcSISP in tobacco (N. nudicaulis) plants conferred enhanced salt tolerance, accompanied by lower cellular Na⁺ contents,¹³⁷ which might be due to the up-regulation of genes associated with Na⁺ exchange and osmotic adjustment. Interestingly, the transgenic tobacco plants displayed higher transcript levels of genes associated with Na⁺ exchange including three members of salt overly sensitive (SOS) pathway (SOS1, SOS2, and SOS3)and three other sodium/hydrogen exchanger (NHX) genes (NHX1, NHX2,andNHX3)under normal conditions.¹³⁷ Although the three hierarchical members of the SOS pathway (SOS1, SOS2, and SOS3) are well-characterized in model plants, particularly A. thaliana,^156–159 however, to our knowledge none of these genes was cloned from citrus. Nevertheless, we believe that the SOS pathway plays a key regulatory role in salt tolerance in citrus (Figure 2(a)).

Figure 2.

Hypothetical model of the potential polyamines- (PAs-) based responses of citrus plants to salinity stress. (a) Integration of PAs with other metabolic pathways in salt tolerance of citrus plants, (b) Activation of salt overly sensitive (SOS) pathway members SOS3, SOS2, and SOS1, respectively. Solid lines with arrowhead signify positive reaction, while dashed lines with blunt-heads indicate negative interaction. For more details and abbreviations, see the main text.

SOS1 encodes a putative plasma membrane Na⁺/H⁺ antiporter,^158,159 SOS2 encodes a serine/threonine protein kinase that is required for salt tolerance,¹⁵⁷ while SOS3 is calcineurin B-like protein (CBL) related to Ca²⁺-binding (calcium sensor).¹⁵⁶ For the activation of the SOS pathway, SOS3 is activated by sensing the changes in cytosolic Ca²⁺ and binding to it (Figure 2(b)).¹⁵⁶ Then, SOS3 interacts with SOS2 and activates it via forming a SOS2-SOS3 protein complex (Figure 2(c)).¹⁵⁷ Later, SOS2 phosphorylates SOS1 (a plasma membrane antiporter) to activate it (Figure 2(d)).^158,159 The active SOS1 controls the exchange of Na⁺ for H⁺ across the plasma membranes and also controls long-distance transportation of Na⁺ within the plant, from roots to stem and leaves.^158,159

In 2015, Alvarez-Gerding et al. genetically-generated a transgenic rootstock C.macrophylla lines that constitutively express the c-repeat-binding factor (CBF3)/dehydration-responsive element-binding 1A (DREB1A) from A.thaliana.¹⁴⁸ CBF3/DREB1A is a well-known salinity tolerance transcription factor. Under salinity stress, the CBF3/DREB1A transgenic lines displayed a higher expression of AtCBF3/DREB1A, as well as the induction of several other genes. For instance, the transgenic lines showed higher expression of ADC2(PAs biosynthetic gene), P5CS (_L-proline biosynthetic gene), and compatible osmolites-associated genes such as galactinol synthase (GOLS), lysine ketoglutarate reductase (LKR/SDH), and raffinose synthase (SIP1).¹⁴⁸ Interestingly, the overexpression of P5CS from moth bean (Vigna aconitifolia) in the citrus rootstock Swingle citrumelo (C. paradisi cv. Duncan x P. trifoliata) induced the expression of antioxidant enzymes genes including cytosolic ascorbate peroxidases (APX1 and APXc1), catalase 2 (CAT2), and cytosolic superoxide dismutase (Cu/ZnSOD2), and also increased the endogenous proline accumulation under water deficit of drought stress.¹⁶⁰ Moreover, the CBF3/DREB1A transgenic lines displayed higher transcripts of dehydrins (COR15A and LEA 4/5) under salt stress.¹⁴⁸ Dehydrins were significantly induced in C. clementina in response to water deficit of drought stress.^75,161 Collectively, we suggest that CBF3/DREB1 could enhance the salinity tolerance in citrus plants via the induction of several genes involved in multiple metabolic pathways putatively belonging to its activation cascade. PAs, _L-proline, compatible osmolytes, and their biosynthetic genes may play a key regulatory role in maintaining the Na⁺ cellular content or ROS homeostasis (Figure 2(a)), however, more investigations are required to understand the molecular and physiological events associated with these compounds.

Role of PAs in citrus response(s) to heavy metals

The term ‘heavy metal’ refers to any relatively high metal or metalloid and has potential toxicity at low concentrations. Heavy metals are environmentally hazardous and serious toxicants carcinogens to animals and humans.¹⁶² Arsenic (As), cadmium (Cd), chromium (Cr), and nickel (Ni) are classified as “category 1 heavy metals; group 1 carcinogens” according to the International Agency for Research on Cancer.¹⁶³ The phytotoxicity of heavy metals is a significant threat to citriculture worldwide. Previously, the effect of various heavy metals on citrus physiology has been investigated including mercury (Hg).¹⁶⁴ cadmium (Cd).^165–169 and chromium (Cr).^170,171 However, the role of PAs in citrus tolerance to heavy metals is poorly studied. To our knowledge, only one study has been recently published that provides strong evidence about the role of PAs in chromium-tolerance in Kinnow mandarin (C. nobilis × C. deliciosa) grafted on the diploid (2x) and double-diploid (4x) rootstocks of P. trifoliata, C. reshni, and C. limonia.¹⁷⁰

Briefly, Shahid et al. (2018a) reported that the free PAs contents, including putrescine, spermidine, spermine, cadaverine, and homospermidine, were significantly higher in chromium-stressed (0.75 mM Cr) plants compared with those grown under normal conditions (0 mM Cr).¹⁷⁰ Likewise, plants exposed to Cr-toxicity had higher levels of soluble-conjugated and insoluble-boundPAs (putrescine, spermidine, spermine, cadaverine, and homospermidine) compared with non-stressed plants. All free, soluble-conjugated and insoluble-boundPAs were induced greatly in the plants grafted on double-diploid rootstocks compared with those grafted on diploid rootstocks.¹⁷⁰ Surprisingly, the chromium-stress significantly reduced the enzymatic activities of four PAs-biosynthetic enzymes including ADC, ODC, SAMDC, and SPDS, while it induced the enzymatic activities of polyamine oxidase (PAO) and diamine oxidase (DAO) (two major PAs-catabolic enzymes).¹⁷⁰

These findings were partially in agreement with the results of Tajti et al. (2018) on wheat (Triticum aestivum) seedling, where putrescine and spermidine were induced by 50 µM cadmium (Cd²⁺) stress.¹⁷² In contrast with these findings, endogenous putrescine and spermidine contents, but not spermine, were decreased by copper- (Cu²⁺) and Cd²⁺-induced oxidative damage in sunflower (Helianthus annuus) leaf discs.¹⁷³ However, pre-treatment with exogenous spermine (1 mM) reverted the negative effects of Cd²⁺ and Cu²⁺ on lipid peroxidation almost to control values.¹⁷³ Similarly, endogenous spermidine and spermine levels were reduced in wheat seedling pre-soaked in H₂O and subjected to Cd²⁺ stress and the pre-soaking in spermidine (2 mM) or spermine (2 mM) mitigated the adverse effects of Cd²⁺ stress (1 mM) and increased the activities of antioxidant enzymes (SOD and CAT) of wheat seedlings.¹⁷⁴ Likewise, the pre-soaking of wheat seeds in aqueous solutions of putrescine (0.5 mM) and spermidine (0.5 mM) modulated the cadmium toxicity and enhancing Cd²⁺-tolerance in wheat seedlings.¹⁷²

Citrus response to cadmium particularly, and heavy metals in general, might be associated with the endogenous phytochelatin (PC) but not glutathione (GSH) content.¹⁶⁸ For instance, the endogenous levels of three phytochelatins (PC-2, PC-3 and PC-4) were significantly increased in Cd²⁺-treated roots of two citrus genotypes included Cleopatra mandarin and Carrizo citrange, whereas both GSH and oxidized GSH were significantly reduced in the Cd²⁺-treated roots.¹⁶⁸ Additionally, the role of PAs in plant response(s) to heavy metals could be correlated with increased endogenous PAs content, accumulation of salicylic acid (SA) and proline, alteration of carbon and nitrogen metabolism, and/or homeostasis of ROS via the regulation of antioxidant enzymes.^170–176 However, the molecular mechanisms behind the PAs-based response of citrus plants against heavy metals remain poorlyunderstood and required further investigations.

Role of PAs in citrus response(s) to water stress

Due to the long lifespan of citrus species, they are subjected to several biotic and abiotic stresses.^177,178 Drought is one of the most intimidating environmental factors for citriculture due to its negative effects on plant growth and development, as well as, productivity.¹⁷⁹ Drought-activated physiological damage results from the imbalance between water uptake by the root system and water loss via shoot system, causing a reversible decline in leaf water relations, membrane stability, gas exchange, and photosynthetic activity, leading to ROS accumulation, lipid peroxidation and membrane injury.^180,181 A systemic response to abiotic stimuli, including drought, is initiated from plant physiological responses by a signaling cascade termed systemic acquired acclimation (SAA).¹⁸² SAA may include changes at the transcriptional, post-transcriptional, proteomic, metabolic changes and/or accumulation,¹⁸³ as well as, ROS, calcium waves, hydraulic waves, and phytohormones, particularly ABA.¹⁸²

In citrus, several metabolites are involved in SAA (Table 5). The PAs-based response is an important strategy used by citrus to tolerate and recover from drought events, as well as, to define the other metabolic-based strategies against water stress. A significant accumulation of both putrescine and spermidine were observed in the leaves of drought-stressed plants of ‘Red Tangerine’ (C. reticulata) ⁵⁷ and ‘Jincheng’ Sweet orange (C. sinensis), ¹¹ but they did not change significantly in water-stressed ‘Owari’ Satsuma orange (C. unshiu).⁴ Moreover, spermine was slightly increased in leaves of ‘Red Tangerine’,⁵⁷ however, it was decreased in other studies on ‘Jincheng’ Sweet orange¹¹ and ‘Owari’ Satsuma orange.⁴ Collectively, these findings indicate that PAs-based response of citrus to water stress is mainly controlled by putrescine and spermidine, but not spermine.

Table 5.

Changes in the endogenous content of major polyamines (PAs), stress-associated amino acids, and stress-associated phytohormones in citrus and its relative species as affected by water stress.

‘Cultivar’/Variety/(species)	Tissue	Treatment/experimental design	Pro a	Arg	JA	ABA	Put	Spd	Spm	Reference
‘Red Tangerine’ (C. reticulata)	Leaves	The role of spermine application (1 mM) in dehydration tolerance after 12 hpt in vitro.	-	-	-	-	▲	▲	▲	57
Tomato (Solanum lycopersicum ‘Zhongshu No. 4′)	Leaves	Overexpression of PtADC gene from P. trifoliata in transgenic tomato plants	-	-	-	-	▲	▲	↑↓	82
Tobacco (Nicotiana nudicaulis) overexpressing PtADC	Leaves	Overexpression of PtADC gene from P. trifoliata in transgenic tobacco plants	-	-	-	-	▲	▲	●	82
Citrumelo CPB 4475 (C. paradisi × P. trifoliata)	Roots	Severe drought stress by transplanting plants to dry perlite for different periods	-	-	▲	▲	-	-	-	184
‘Owari’ Satsuma orange (C. unshiu)	Leaves	Dropping the soil moisture content to 20% of the water holding capacity for 5 days	▲	▼	-	-	●	●	▼	4
Tobacco (N. nudicaulis) overexpressing FcWRKY70	Leaves	Overexpression of FcWRKY70 from ‘Meiwa’ Kumquat (Fortunella crassifolia)	-	-	-	-	▲	-	-	185
Cleopatra mandarin (C. reticulata),	Leaves	Progressive drought was imposed by transplanting plants to perlite, ceasing watering and allowing the substrate to dry for a period of 14 days.	▲	-	▲	▲	-	-	-	186
Carrizo citrange (C. sinensis × P. trifoliata)	Leaves	Progressive drought was imposed by transplanting plants to perlite, ceasing watering and allowing the substrate to dry for a period of 14 days.	▲	-	▲	▲	-	-	-	186
‘Jincheng’ Sweet orange (C. sinensis)	Leaves	The seedlings were dehydrated in vitro at 25°C	-	-	-	-	▲	▲	▼	11
Rangpur Lime (C. limonia) b	Leaves & Roots	Either ungrafted or grafted subjected to distinct watering regimes (well-watered vs. severe drought)	-	-	-	▲	-	-	-	180
‘Sunki Maravilha’ mandarin (C. sunki) b	Leaves & Roots	Either ungrafted or grafted subjected to distinct watering regimes (well-watered vs. severe drought)	-	-	-	▲	-	-	-	180
‘Valencia’ Sweet Orange (C. sinensis) on C. limonia	Leaves & Roots	Responses of Valencia scion grafted on two rootstocks, Rangpur Lime (C. limonia) and ‘Sunki Maravilha’ mandarin (C. sunki), during three successive periods of water deficit	-	-	-	▲	-	-	-	187
‘Havana’ Tobacco (N. tabacum) overexpressing CsNCED3	Leaves	Overexpression of CsNCED3 from ‘Rangpur’ lime (C. limonia)	-	-	-	▲	-	-	-	188

^a(▲) signifies for increased levels, (▼) signifies for decreased levels, (●) signifies for no significant change, (↑↓) signifies for varied responses, and (˗) signifies for not-reported compounds. Abbreviations: Pro: _L-proline, Arg:_L-Arginine, JA: jasmonic acid, ABA: abscisic acid, Put: putrescine, Spd: spermidine, and Spm: spermine.

^bThe two Citrus rootstocks (Rangpur lime and ‘Sunki Maravilha’ mandarin) either ungrafted or grafted with their reciprocal graft combinations or with shoot scions of ‘Valencia’ Sweet Orange (C. sinensis) and Tahiti acid lime (C. latifolia) subjected to distinct watering regimes (well-watered vs. severe drought)

Although several genes have been genetically transformed from/to citrus or its relatives and their transgenic plants were more tolerant to water stress (Table 6), however, very few of them had higher PAs levels. For instance, overexpression of an arginine decarboxylase (PtADC) gene, isolated from Poncirus trifoliata, in tomato (Solanum lycopersicum ‘Zhongshu No. 4′) and tobacco (N. nudicaulis) showed elevated putrescine and spermidine levels in the transgenic lines.⁸² Likewise, overexpression of FcWRKY70 from ‘Meiwa’ Kumquat (Fortunella crassifolia) in tobacco (N. nudicaulis) plants increased endogenous putrescine content and up-regulated the ADC genes (NtADC1 and NtADC2) in tobacco transgenic plants compared with wild-type plants.¹⁸⁵ Interestingly, challenging the transgenic lines with dehydration for 60 min led to significant putrescine accumulation in the tested genotypes.¹⁸⁵ Additionally, the transcript levels of both ADC genes (NtADC1 and NtADC2) were induced after dehydration treatment, but they were pronouncedly higher in the transgenic lines than in WT.¹⁸⁵

Table 6.

Water stress response-associated genes that have been genetically transformed from/to citrus or its relatives.

Genes	GenBank accession No. a	Source	‘Cultivar’/Variety/(species)	References
P5CS	-	Moth bean (Vigna aconitifolia)	‘Carrizo’ citrange (C. sinensis × P. trifoliata)	178
PtrABF	HM171703.1	Trifoliate Orange (P. trifoliata)	Tobacco (Nicotiana nudicaulis)	189
CuLea5 (LeaP600)	DQ424891.1	Satsuma mandarin (C. unshiu)	Arabidopsis thaliana ecotype ‘Columbia’	190
PtrMAPK	-	Trifoliate orange (P. trifoliata)	Tobacco (N. tabacum)	191
PtrADC	HQ008237	Trifoliate Orange (P. trifoliata)	Tomato (Solanum lycopersicum ‘Zhongshu No. 4′)	82
PtrADC	HQ008237	Trifoliate Orange (P. trifoliata)	Tobacco (Nicotiana nudicaulis)	82
CpPSY	-	‘Flame’ Grapefruit (Citrus paradisi)	‘Havana’ tobacco (N. tabacum)	192
P5CS	-	Moth bean (Vigna aconitifolia)	‘Duncan’ Swingle citrumelo (C. paradisi x P. trifoliata)	160,193
CsGSTUs	-	Sweet orange (C. sinensis)	‘Wisconsin 38ʹ tobacco (N. tabacum)	194
FcWRKY70	-	‘Meiwa’ Kumquat (Fortunella crassifolia)	Tobacco (N. nudicaulis)	185
FcWRKY70	-	‘Meiwa’ Kumquat (F. crassifolia)	Lemon (C. lemon)	185
PtrABF	HM171703.1	Trifoliate Orange (P. trifoliata)	Trifoliate orange (P. trifoliata)	195
CsTIP2	-	‘Rangpur’ Lime (C. limonia)	‘Havana’ tobacco (N. tabacum)	196
CsNCED3 b	-	‘Rangpur’ Lime (C. limonia)	‘Havana’ tobacco (N. tabacum)	188

^a(˗) Signifies for not-reported accession numbers.

^bThis gene corresponding to the identical sequence of locus ‘Ciclev10019364 m’ in the citrus reference genome.

Additionally, ectopic expression of ABA-responsive element (ABRE)-binding factor from P. trifoliata (PtrABF) in transgenic tobacco plants enhanced tolerance to both dehydration and drought stresses through scavenging ROS and modulating the gene expression of PAs-biosynthetic genes (NtADC1, NtADC2, and NtSAMDC) and stress-responsive genes (NtLEA5, NtERD10 C, NtCDPK2, and NtDREB).¹⁸⁹ Similarly, overexpression of the same gene in trifoliate orange (P. trifoliata) reduced the stomatal density, maintained ROS homeostasis, up-regulated PtrADC, and induced nine stress-associated transcription factors of the NACs, WRKYs, and bZIP families, and seven stress-associated proteins including LEA4-5 and dehydration-induced protein RD22-like.¹⁹⁵ Dehydration for 90 min increased the transcript levels of PtrADC in both WT and transgenic lines; however, the induction was greater in the transgenic lines than in wild-type plants.¹⁹⁵ In all these studies and under water stress conditions, the transgenic lines exhibited improved dehydration and drought tolerance,^{82,185,189,195} decreased ROS accumulation,^82,189,195 and up-regulated stress-responsive genes.^{82,185,189,195} In addition, we believe that PAs-based response of citrus to water stress is integrated with other metabolic pathways (Table 5) and water stress response-associated genes/proteins (Table 7). For example, ABA plays a key role in the coordinated regulation of the proline and polyamine biosynthetic pathways. For example, both ABA and osmotic stress, caused by polyethylene glycol altered the proline metabolism differently, however, ABA treatment resulted in higher proline accumulation in wheat seedlings.¹⁹⁸ Remarkably, proline metabolism was partly regulated independently and not in an antagonistic manner from PAs biosynthesis.¹⁹⁸ Together, these findings support our suggestion that PAs-based response of citrus to water stress is mainly controlled by putrescine and spermidine, but not spermine and regulated by other stress-associated metabolites, particularly ABA and proline.

Table 7.

Water stress response-associated genes that have been cloned from citrus or its relatives.

Genes	GenBank accession No. a	‘Cultivar’/Variety/(species)	References
CsLea5	-	Sweet orange (C. sinensis)	140
CsNCED1	DQ028471	‘Navelate’ navel orange (C. sinensis)	197
CsNCED2	DQ028472	‘Navelate’ navel orange (C. sinensis)	197
PtrRING‐H2	DQ158860	‘Rubidoux’ Trifoliate Orange (P. trifoliata)	76
PtrABF	HM171703.1	Trifoliate Orange (P. trifoliata)	189
CsSAMDC	FJ496345	‘Newhall’ Navel orange (C. sinensis)	79
CsNAC1	CX306190	‘Cleopatra’ Mandarin (C. reshni)	80
PtrMAPK	-	Trifoliate orange (P. trifoliata)	191
PtrADC	HQ008237	Trifoliate Orange (P. trifoliata)	82
PtrWRKY2	JX236033.2	‘Rubidoux’ Trifoliate Orange (P. trifoliata)	84
PtrWRKY1	CX065984	‘Rubidoux’ Trifoliate Orange (P. trifoliata)	86
CgWRKY1	-	‘Rein King’ Pummelo (C. grandis)	86
CrNCED1	DQ028471.1	‘Cleopatra’ mandarin (C. reshni)	147
PtrABF	HM171703.1	Trifoliate Orange (P. trifoliata)	195
Four CsWRKYs(Cs30, Cs40, Cs41, and Cs48)	-	Sweet orange (C. sinensis)	152
Four CcWRKYs (Cc53, Cc54, Cc60, and Cc67)	-	Clementine (C. clementina)	152
PtrZPT2-1	KC820894	Trifoliate orange (P. trifoliata)	153

^a(˗) Signifies for not-reported accession numbers.

PAs-Based Defensive Mechanisms Against Biotic Stress in Citrus

It has been suggested that plant metabolites are responsible for securing the energy requirements for plant defensive responses against biotic stress, particularly phytopathogens,^199,200 however, this is not the main role. Plant metabolites might contribute directly to plant defense responses, or they could be required to nutritional needs for pathogen growth and multiplication. Previously, we described the chemical composition of citrus phloem sap because it contains all the required nutrients for growth and multiplication of several citrus pathogens such as Candidatus Liberibacter spp., the causal agents of Huanglongbing (aka citrus greening).²⁰¹ GC-MS analysis revealed that Citrus phloem-sap contains many proteinogenic amino acids (PAAs; such as proline, tryptophan, tyrosine, leucine, isoleucine, and valine), sugars (sucrose, glucose, fructose, and inositol), fatty acids (palmitic, oleic, and stearic acids), and organic acids (succinic, fumaric, malic, maleic, threonic, citric, and quinic acids).²⁰¹ Proline, the most abundant amino acid, while γ–aminobutyric acid (GABA) was the only NPAA detected in citrus phloem sap.²⁰¹

Citrus plants are threatening by several biotic stressors such as viral, bacterial, and fungal diseases, in addition to many nematodes, phytoplasmas, spiroplasmas, viroids, and graft-transmissible pathogens. All these pathogens can severely influence many biochemical and metabolic alterations that play a vital and complex role in citrus defensive mechanisms.^{1–8,12,202,203} The metabolic changes including the alteration of volatile organic compounds,²⁰³ amino acids,^1–6 organic acids,^1,3,6 fatty acids,^2,6 phytohormones,^7,8 polyamines (PAs)^4,12 and other secondary metabolites. However, it is still ambiguous whether these changes are a plant metabolic response(s), or if they are to benefit citrus pathogens or both. In addition, the contribution of PAs and NPAAs pathways in citrus defensive mechanisms against biotic stress are poorly studied and not fully understood.

Role of PAs in citrus response(s) to virus and viroid infection

Citrus species could be infected by many viruses, most of them are asymptomatic until a susceptible scion and/or rootstock is exist.²⁰⁴ Citrus viruses including, but not limited to, Citrus leaf blotch virus (Citrivirus), Citrus leprosis (Cilevirus/Higrevirus), Citrus chlorotic dwarf (Geminiviridae), Citrus psorosis virus (CPsV; Ophiovirus), and Citrus tristeza virus (CTV; Closterovirus).²⁰⁴ In addition, citrus plants could be attacked by a large number of true viroid species; including Citrus Exocortis Viroid (CEVd), Hop Stunt Viroid (HSVd), Citrus Bent Leaf Viroid (CBLVd), Citrus Dwarfing Viroid (CDVd), and Citrus Bark Cracking Viroid (CBCVd).²⁰⁵ Two additional viroids, Citrus Viroid VI (CVd-VI) and Citrus Viroid V (CVd-V) have also been reported.^206–208 To our knowledge, the role of PAs in citrus-virus or citrus-viroid pathosystems is not studied yet.

Many attempts have been made to create transgenic resistant plants for CTV and CPsV (Table 8), via introducing the coat protein (CP) genes of CTV^{209–211,213,214,222,227} or the coat protein (CPsV-54 K) gene and viral 24 k-derived hairpin of CPsV (CPsV-ihp24 K; apolypeptide of unknown function)^220,221,225 to virus-sensitive cultivars. However, the role of PAs in citrus response to CTV or CPsV is not reported in these transgenic plants. On the other hand, very few reports about the correlation between CEVd and PAs in tomato (Lycopersicon esculentum) and purple passion (Gynura aurantiaca) plants were found, but not in citrus.^228,229 Recently, it has been proved that the nonproteinogenic amino acid, γ-aminobutyric acid (GABA) is associated with tomato response against CEVdafter the treatment with the plant defense inducer, benzothiadiazole (BTH).²³⁰

Table 8.

Viral disease tolerance genes that have been cloned/genetically transformed from/to citrus or its related genera.

Gene	‘Cultivar’/Variety/(species)	References
CTV-CP gene	Sour orange (Citrus aurantium)	209
CTV-CP gene	Mexican lime (C. aurantifolia)	210
CTV-CP gene	Sour orange (C. aurantium)	211
CTV-uncp	‘Rio Red’ grapefruit (C. paradisi)	212
gna	‘Rio Red’ grapefruit (C. paradisi)	212
CTV-p25 CP	Mexican lime (C. aurantifolia)	213
CTV-CP gene	Grapefruit (C. Paradisi)	214
CTV-p23	Mexican lime (C. aurantifolia)	215,216
CTV-RdRp	‘Duncan’ grapefruit (C. paradisi)	217
CTV-392 CTV-p23 hairpin	Grapefruit (C. paradisi) Grapefruit (C. paradisi)	218 218
scFv-3DF1	Mexican lime (C. aurantifolia)	219
scFv-3CA5	Mexican lime (C. aurantifolia)	219
CPsV-CP	Sweet orange (C. sinensis)	220
CPsV-24 K gene CPsV-48 K gene CPsV-54 Kgene	Sweet orange (C. sinensis) Sweet orange (C. sinensis) Sweet orange (C. sinensis)	221 221 221
pCTV-CP pCTV-dsCP pCTV-CS	Sweet orange (C. sinensis) Sweet orange (C. sinensis) Sweet orange (C. sinensis)	222 222 222
CTV-p20 hairpin RNA	sour orange (C. aurantium)	223
CTV-p23	Mexican lime (C. aurantifolia)	224
CPsV-ihp24 K	‘Pineapple’ sweet orange (C. sinensis)	225
hp2520	‘Jincheng’ sweet orange (C. sinensis)	226
hp252023	‘Jincheng’ sweet orange (C. sinensis)	226
CTV-CP gene	Sour orange (C. aurantium)	227

Abbreviations:

Citruspsorosis virus (CPsV)

Citrus tristeza virus(CTV)

CTV-CP: Citrus tristeza viruscoat protein

CTV-uncp:Citrus tristeza virus untranslatable coat protein gene

Gna: Galanthus nivalis agglutinin gene (a plant-derived insecticidal gene)

CTV-RdRp: Citrus tristeza virus RNA-dependent RNA polymerase gene

scFv: Single-chain variable fragment antibodies

pCTV-CP: Containing the 25 kDa major capsid protein gene (CTV-CP),

pCTV-dsCP: Containing the same CTV-CP gene in an intron-spliced hairpin construct

pCTV-CS: Containing a 559 nt conserved region of the CTV genome

ihp24 K: Citruspsorosis virus (CPsV) viral 24 k-derived hairpin

hp252023: Hairpin RNA targeting p25, p20, and p23

hp2520: Hairpin RNA targeting p25 and p20

CPsV-24 K gene: A 24‐kDa polypeptide of unknown function (24 K) from CPsV

CPsV-48 K gene: A polypeptide of 54.7 kDa of unknown function (54 K) from CPsV

CPsV-54K gene: The coat protein (CP) of 48.6 kDa (48 K protein) from CPsV

CEVd infection stimulated the ethylene production via the activation of 1-Aminocyclopropane-1-carboxylic acid (ACC) synthesis(Figure 3(a)) ²³¹ and enhanced the production of pathogenesis-related (PR) proteins ²²⁸ in tomato leaves when systemic symptoms appeared. On the other hand, the CEVd-infected tomato plants displayed a marked decrease in endogenous putrescine content, but no significant changes in either spermidine or spermine content.²²⁸ The reduction of free putrescine content caused by CEVd infection was correlated with the reduction in ODC activity, but not ADC activity.²²⁹ In addition, treatment with aminoethoxyvinylglycine (AVG; an ethylene biosynthesis inhibitor) or norbornadiene (an ethylene action inhibitor) frustrated the reduction in ODC activity and putrescine content without no significant effect on spermidine or spermine content.^228,229 Taken together, these findings suggest that ethylene is negatively regulating the putrescine during the viroid-plant interaction via interfering with the ODC activity (Figure 3(a)).

Figure 3.

Hypothetical model of the potential polyamines- (PAs-) based responses of citrus plants against viroid diseases. (a) Integration of PAs with ethylene in citrus response to the infection with Citrus Exocortis Viroid (CEVd), and (b) The interaction between PAs, ethylene, and salicylic acid during the CEVd infection. Solid lines with arrowhead signify positive reaction, dashed lines with blunt-heads indicate negative interaction, while round-dotted lines with arrows represent hypothetical mechanisms or uncharacterized elements. For more details and abbreviations, see the main text.

Moreover, salicylic acid (SA) is another phytohormone that is involved in the basal resistance of tomato plants to CEVd infection.²³² CEVd infection induced the accumulation of SA in the ‘Money Maker’ tomato cultivar, but not in tomato plants expressing the NahG transgene (a mutant prevents the accumulation of endogenous SA).²³² In addition, CEVd infection increased the levels of ethylene and putrescine conjugates (caffeoyl-putrescine, and feruloyl-putrescine) in NahG plants earlier and greater than in Money Maker plants.²³² These findings suggest that SA also might function as a negative regulator of both ethylene and putrescine conjugates during the viroid-plant interaction (Figure 3(b)).

On the other hand, the viral infection with CTV induced the putrescine accumulation (approximately 6-folds) in the phloem sap of CTV-infected C. macrophylla plants.⁴⁴ Likewise, PAs were accumulated in hypersensitive, but not susceptible, tobacco plants upon the infection with Tobacco Mosaic Virus (TMV).²³³ Although the induction of PAs against TMV appears to function independently from SA to activate acidic PR proteins and resistance against virus infection in tobacco plants,²³⁴ several transcriptomic and proteomic studies revealed that phytohormones might be implicated in citrus response against CTV.^235–239 In citrus, PAs-based defensive mechanisms against CTV might be integrated with phytohormone groups (Figure 4) such as abscisic acid (ABA), ^236,238,240 auxins,^239,240 brassinosteroids (BRs)and cytokinins (CKs),²⁴⁰ ethylene,^{235,237–240} gibberellins (GA), ^238–240 jasmonates (JA), ^237,240 andsalicylic acid (SA).^235,238

Figure 4.

Hypothetical model of the potential polyamines- (PAs-) based responses of citrus plants against viral diseases. (a) Integration of PAs with seven phytohormone pathways in citrus response to the infection with Citrus tristeza virus (CTV), and (b) The interaction between PAs, auxins, cytokinins (CKs), and methyl jasmonates (meJA) during the CTV infection. Solid lines with arrowhead signify positive reaction, dashed lines with blunt-heads indicate negative interaction, while round-dotted lines with arrows represent hypothetical mechanisms or uncharacterized elements. For more details and abbreviations, see the main text.

Four instances, four ABA-related genes were up-regulated in CTV-infected sweet orange (C. sinensis cv. ‘Jincheng’) leaves compared with mock-inoculated ones using microarrays.²⁴⁰ These differentially expressed genes (DEGs) including 9-cis-epoxycarotenoid dioxygenase 4 and 5 (NCED4 and NCED5), GRAM (glycine, arginine, alanine and methionine motif) domain-containing protein/ABA-responsive protein-related, and HVA22-like protein F (HVA22 F).²⁴⁰ Likewise, NCED4 wasup-regulated in sweet orange trees affected by the CTV-B6 severe strain, but not the mild strain (CTV-B2).²³⁸ Moreover, the ABA-stress ripening-related protein was increased in CTV-infected sweet orange (C. sinensis cv. ‘Westin’) plants.²³⁶ Interestingly, NCED3 was up-regulated was induced in spermine-treated trifoliate orange seedlings,⁵⁸ which supports the idea that PAs-based defensive mechanisms against CTV might be integrated with ABA and its related genes (Figure 4(a)).

Additionally, auxins metabolism and signaling-related genes were differentially regulated upon CTV infection.^239,240 Five genes of auxin-responsive family protein were down-regulated in CTV-infected sweet orange leaves compared with mock-inoculated plants.²⁴⁰ Nevertheless,CTV-infection induced the expression of the auxin-responsive GH3 protein (GH3.1)²⁴⁰ and auxin signaling F-box protein 5 (AFB5).²³⁹ GH3 gene is involved in the conversion of free active auxins to inactive auxin-amide conjugates via conjugates amino acids to indole-3-acetic acid (IAA).^241,242 Together, these findings suggest that CTV infection might decrease the free active auxins level viathe reduction of auxin-responsive family proteins and the induction of GH3(Figure 4(a)). On the other hand, four CKs-related DEGs were down-regulated in CTV-infected sweet orange leaves compared with mock-inoculated plants.²⁴⁰ These DEGs include cytokinin oxidase 5 (CKX5), isopentenyltransferase 3 (IPT3), and two isopentenyltransferases 5 (IPT5).²⁴⁰ The CKs reduction might be a defense mechanism to suppress the development of the symptoms. In Arabidopsis-Rhodococcus fascians interaction, CKs were involved in symptom development via activation of ADC and induction of PAs accumulation.²⁴³

Furthermore, four BRs-related DEGs were increased in CTV-infected sweet orange leaves compared with mock-inoculated plants.²⁴⁰ These DEGs include steroid sulfotransferase (ST), FACKEL (FK; encodes a protein with sequence similarity to sterol reductase family), and two squalene epoxidases 3 (SQE3).²⁴⁰ SQE3 is a rate-limiting enzyme in the sterol biosynthesis pathway and might play a key role in citrus defense against CTV. ST gene also might be involved in citrus defense response against CTV, since it found to be induced by SA.²⁴⁴ SA-defense pathway mediated the resistance of sour orange (C. aurantium) rootstock to CTV.²⁴⁵ Additionally, indirect evidence for the alteration of SA in Mexican lime plants infected by the CTV severe strain (CTV-T305),²³⁵ since S-adenosyl-methionine salicylic acid carboxyl methyltransferase (SAMT) gene was up-regulated upon CTV-infection. SAMT catalyzes the formation of methylsalicylate (meSA) from SA and induces in systemic acquired resistance (SAR).²⁴⁶ The accumulation of meSA was reported previously from TMV-infected tobacco plants.²⁴⁷ In addition, exogenous application of SA derivative (acetylsalicylic acid; aka aspirin) induced resistance to TMV in tobacco.²⁴⁸ Interestingly, SA and its derivative, meSA, enhanced the accumulation of PAs in other plant species such as maize,²⁴⁹ tobacco,²⁵⁰ and tomato²⁵¹ via the induction of ADC and/or ODC. Taken together, these findings suggest that PAs-based defensive mechanisms against CTV are mediated by both BRs and SA.

The importance of JA and ethylene pathways as alternative signals in plant defense was magnified²⁵² due to their association with PAs metabolism.²⁵³ For example, ACC oxidase (an enzyme that catalyzes ACC to ethylene) was up-regulated in CTV-infected Mexican lime plants.²³⁵ Likewise, 29 ethylene-related genes were differentially expressed in CTV-infected sweet orange (C. sinensis cv. ‘Jincheng’) leaves.²⁴⁰ The expression of ethylene-responsive element binding factor 1 (ERF1; an ethylene signal transduction) and two of senescence-related gene 1 (SRG1) were also increased in sweet orange trees infected with a severe strain of CTV (CTV-B6).²³⁷

Moreover, nine JA-related genes were differentially expressed in CTV-infected plants.²⁴⁰ JA carboxyl methyltransferase (JMT) was highly up-regulated upon CTV-infection.^237,240 Additionally, several JA-biosynthetic genes were expressed at higher levels in CTV-infected plants including two fatty acid desaturase (FAD3and FAD6),hydroperoxide lyase 1 (HPL1), twoallene oxide synthase (AOS), twoallene oxide cyclase 4 (AOC4),and two 12-oxophytodienoate reductase 2 (OPR2).^237,240 The induction of JA-biosynthetic genes suggesting that these DEGs might play a key role in citrus defense against CTV through the enhancement of endogenous JA content, which could stimulate the PAs-based defensive mechanisms against CTV. Previous studies showed that the exogenous application of methyl jasmonates (meJA) enhanced the accumulation of conjugated PAs via the up-regulation of gene expression and enzymatic activities of PAs-biosynthetic genes (ADC, ODC, SAMDC).²⁵³ However, meJA-induced PAs and their defense-related process are regulated by other phytohormone groups. For example, auxins (such as IAA) synergistically strengthen the meJA-induced PAs, while N⁶-benzyladenine (a synthetic cytokinin) counteracts it, whereas it appears to function independently of ethylene ²⁵³ (Figure 4(b)).

Furthermore, PAs-based defensive mechanisms against CTV might be integrated with the detoxification of ROS (Figure 4(a)). The CTV infection induced the generation of ROS in the infected cells, which subsequently reprogramming both primary and secondary metabolic pathways ²⁴⁰ and influencing the antioxidant enzymes activity.^236,254 For instance, the transcripts and enzymatic activity of superoxide dismutase (SOD) and ascorbate peroxidase (APX) were increased in sweet orange (C. sinensis cv. ‘Westin’) plants upon the CTV infection, while the activity of guaiacol peroxidase (GPX) did not change significantly.²³⁶ However, The enzymatic activities of SOD and polyphenol oxidase (PPO) were significantly reduced in Mexican lime (C. aurantifolia) plants after ten years of exposure to CTV infection.²⁵⁴ These findings suggest that the antioxidant defense mechanisms might be suppressed by CTV because of continuous and long‐term stress.

In conclusion, we suggest that citrus plants relay on multiple defensive mechanisms against viral infection, particularly CTV. The activation of these defensive mechanisms varies according to virus strain, host genotype, time of exposure to the infection, and environmental conditions. These defensive mechanisms include a combination of PAs-based defensive mechanisms, different phytohormone groups, and antioxidant enzymatic activities. However, we believe that more investigations are required to reconnoiter the molecular mechanisms behind the PAs-phytohormones-ROS crosstalk during the citrus-virus interaction.

Role of PAs in citrus response(s) to fungal and oomycetes infection

Citrus and its relative species are attacked by many fungal and oomycetes phytopathogens that can infect citrus fruits and foliage. The fungal diseases ofcitrus includingAlternaria brown spot (Alternaria alternate; synonyms: A. citri and A. alternate pv. citri), greasy spot (Mycosphaerella citri), melanose (Diaporthe citri; synonyms: Phomopsis citri), scab diseases (Elsinoë fawcettii; synonym: Sphaceloma fawcettii), black spot (Phyllosticta citricarpa; synonyms: Guignardia citricarpa and Phoma citricarpa), post-bloom fruit drop (Colletotrichum acutatum;synonyms: Glomerella acutata and C. abscissum),Pseudocercosperafruit and leaf spot (Pseudocercospora angolensis;synonyms: Phaeoramularia angolensis), blue mold (Penicillium italicum) and green mold (P. digitatum).^255,256 In addition, the oomycetous pathogen Phytophthora citrophthora (synonyms: P. citricola and P. parasitica) causes several diseases to citrus plants including foot rot, gummosis, and root rot diseases, and also can infect citrus fruits causing brown rot disease.^255,256

Unfortunately, the available studies about the role of PAs in citrus response(s) to the fungal and oomycetes diseases are very limited. As we mentioned above, citrus plants have two routes for the biosynthesis of putrescine, one from arginine via agmatine involving ADC and the other one from ornithine involving ODC. However, fungal phytopathogen and fungi, in general, have only one PAs biosynthesis pathway through.²⁵⁷ The ODC activity plays a key role in the growth of^257–259 and pathogenicity^257,260 of phytopathogenic fungi (Figure 5(a)). The addition of DL-α-Difluoromethylornithine (DFMO; an inhibitor of ODC) controlled the growth of several phytopathogenic fungi in vitro ^257–259 and also prevented the infection by uredospores of the bean rust fungus, Uromyces phaseoli on bean plants Phaseolous vulgaris.²⁵⁷ Interestingly, the transcriptome analysis of the oomycetous pathogen Phytophthora parasitica, the causal agent of the citrus gummosis disease, revealed that only one singleton read was found in the PP/CitEST database that sharing sequence similarity with the ODC gene of Mucor circinelloides f. lusitanicus.²⁶⁰This enzyme might be an important pathogenicity factor,²⁶⁰ and probably could be a target for chemical control because its importance in virulence and growth, however, these hypotheses are not characterized yet and required further investigations.

Figure 5.

Hypothetical model of the potential polyamines- (PAs-) based responses of citrus plants against fungal diseases. (a) The role of PAs in the pathogenicity of fungi and fungi-like oomycetes and (b) The role of PAs in resistance to fungal diseases of citrus. Solid lines with arrowhead signify positive reaction, dashed lines with blunt-heads indicate negative interaction, while round-dotted lines with arrows represent hypothetical mechanisms or uncharacterized elements. For more details and abbreviations, see the main text.

Furthermore, the exogenous application of 24-epibrassinolide (EBR) on Satsuma mandarin (C. unshiu) fruits significantly enhanced the endogenous ornithine content, which was correlated with a reduced disease incidence of postharvest rots.²⁶¹ Similarly, theheat treatment significantly increased the ornithine content which was associated with reduced disease incidence, lesion diameter, and improved the disease resistance againstP. italicum on mature ‘Kamei’ Satsuma mandarin (C. unshiu) fruits during postharvest storage.²⁶² In both studies, the accumulation of ornithine was associated with higher enzymatic activities of antioxidant enzymes such as CAT, POD, and SOD.^261,262 Collectively, these findings suggest that ornithine might be involved in postharvest disease resistance via regulation of oxidative stress tolerance in citrus plants(Figure 5(b)).

Moreover, transcriptomic profiling of ‘Powell’ orange (C. sinensis) fruit reveal that several genes involved in the PAs pathway were up-regulated after 60-hours of infection with P. digitatum.²⁶³ Thesegenes included the major PAs biosynthetic genes (ADC and ODC), putative acetylornithine aminotransferase, arginine biosynthesis protein (argJ1), putative arginine/serine-rich splicing factor, polyamine oxidase 1, and putrescine-binding periplasmic protein-related.²⁶³ In the same manner, the endogenous PAs levels, particularly putrescine, were induced in orange fruits after 60-hours of the treatment with the extract of the cell-free culture of the biocontrol yeast Kloeckera apiculata.²⁶⁴ K. apiculata was isolated from the epiphytes of citrus roots²⁶⁵ and showed strong antagonistic activities against both P. digitatum²⁶⁶ and P. italicum.^265,267 Together, these findings suggest that the biological control of postharvest fungal pathogens in citrus using K. apiculata is associated with the up-regulation of PAs biosynthesis to enhance citrus responses (Figure 5(b)).

Additionally, at the cell wall level, PAs might help in the prevention of the pathogen entrance by supporting the formation of physical barriers through lignification and suberization.²⁶⁸ For instance, the heat treatment significantly induced the ornithine accumulation in ‘Kamei’ Satsuma mandarin (C. unshiu) pericarp, which was associated with higher lignin content and lower disease incidence of P. italicum.²⁶² Likewise, the lignin content was increased in discs of orange peel after 60-hours of the treatment with K. apiculata or its cell-free extract²⁶⁴ which might be involved in enhancement of fruit resistibility in response to postharvest fungal infection. In conclusion, PAs play a key role in citrus response(s) to fungal infections at multiple levels including regulation of oxidative stress and/or supporting the formation of physical barriers through lignification(Figure 5(b)).

Role of PAs in citrus response(s) to Bacterial infection

Citrus and its relative species are threatened by several bacterial pathogens that cause considerable economic losses and a severe reduction in fruit quality worldwide.²⁶⁹ The citrus bacterial disease including bacterial spot(Xanthomonas alfalfae subsp. citrumelonis), black pit and blast (Pseudomonas syringae), citrus canker (Xanthomonas citri subsp. citri [Xcc]), citrus variegated chlorosis (CVC; Xylella fastidiosa), Huanglongbing (HLB; Candidatus Liberibacter spp.), and stubborn disease (Spiroplasma citri).^255,256,269 Among these diseases, citrus canker, CVC, and HLB are the most economically important bacterial disease of citrus, therefore, these three diseases will be our focus in the next paragraphs.

PAs-based defense responses against citrus canker

Citrus canker is a destructive citrus bacterial disease worldwide, that caused by a Gram-negative obligately aerobic, straight rods Gammaproteobacterium belonging to genus Xanthomonas.^270,271 Citrus canker is affecting several genera from the Rutaceae family such as Citrus spp., Fortunella spp., and Poncirus spp., including the economically important cultivated hosts such as orange, grapefruit, lime, lemon, pomelo, and citrus rootstock.²⁷¹ Citrus canker disease is caused by two groups of phylogenetically distinct Xanthomonas strains. The first group (Asiatic group) named Xanthomonas citri (synonyms: X. campestris pv. citri pathotype A and X. axonopodis pv. citri), while the second phylogenetically distinct group (South American group) named X. campestris pv. aurantifolii (synonyms: X. campestris pv. citri and X. axonopodis pv. aurantifolii).^270,271 These groups cause identical symptoms using, almost, similar common pathogenicity factors/mechanisms,however, both groups differ in their natural occurrence and host range.^269,271

Although enormous work has been focused on creating transgenic citrus plants for canker resistance (Table 9), only one trail was done to target a PAs biosynthetic gene.²⁹⁴ Fu et al. (2011a) showed that the overexpression of apple (Malus sylvestris) spermidine synthase gene (MdSPDS1) in ‘Anliucheng’ sweet orange (C. sinensis) increased the cellular endogenous PAs contents but reduced the susceptibility of transgenic plants to X. axonopodis pv. citri (Xac), the causal agent of citrus canker.²⁹⁴ Upon Xac infection, the transgenic line TG9 had significantly higher spermine levels, PAO activity, antioxidant enzymatic activities (SOD and CAT) and induced H₂O₂ accumulation.²⁹⁴ The accumulation of H₂O₂ could be derived from the PAO-mediated oxidation of spermidine or spermine^304,305 to trigger the hypersensitive response (HR) or to activate the defense-related genes such as chitinase, cystatin-like protein, pathogenesis-related protein 4A (PR4A) and AOS.^294,295 In addition, the genetic modification of PAs biosynthesis pathway may cause transcriptional reprogramming in the transgenic plants, which might be regulated by spermine as it was previously proposed as a signal molecule.^234,306 For example, 13 transcription factors (TFs) were upregulated in the canker-resistant transgenic line TG9 that had a higher spermine level.²⁹⁵ The transcriptional up-regulation toward the disease tolerance including AP2 domain-containing transcription factor, three AP2/ERF domain-containing, BT4 (BTB and TAZ domain protein 4) protein, chromatin remodeling complex subunit, FLC-like 1 splice variant 4, guanylyl cyclase, MADS, MYB, NAC, and two unnamed putative TFs²⁹⁵ (Figure 6).

Table 9.

Genetic transformation, transient expression, or over-expression in citrus and its related genera of genes that have been conferred resistance against citrus canker disease.

Genes	Source	‘Cultivar’/Variety/(species)	References
Attacin A	Tricloplusia ni	`Hamlin’ Sweet Orange(C. sinensis)	272
	T. ni T. ni T. ni	`Valencia’ Sweet Orange(C. sinensis) `Natal’ Sweet Orange(C. sinensis) `Pera’ Sweet Orange(C. sinensis)	273 273 273
Xa21	Wild rice (Oryza longistaminata)	‘Early Gold’ sweet orange (C. sinensis)	274
	Wild rice (O. longistaminata)	`Hamlin’ Sweet Orange(C. sinensis)	275-277
	Wild rice (O. longistaminata) Wild rice (O. longistaminata) Wild rice (O. longistaminata)	`Valencia’ Sweet Orange(C. sinensis) `Natal’ Sweet Orange(C. sinensis) `Pera’ Sweet Orange(C. sinensis)	277 277 277
	Wild rice (O. longistaminata)	‘Anliucheng’ sweet orange (C. sinensis)	278
FLS2 CpFLS2-2 CrFLS2-2 FmFLS2-1	Tobacco (Nicotiana benthamiana) Tobacco (N. benthamiana)	`Hamlin’ Sweet Orange(C. sinensis) ‘Carrizo’ citrange (C. sinensis × P. trifoliata)	279 279
	‘Duncan’ grapefruit (C. paradisi) ‘Sun Chu Sha’ mandarin (C. reticulata) ‘Nagami’ kumquat (Fortunella margarita)	‘Duncan’ grapefruit (C. paradisi) ‘Duncan’ grapefruit (C. paradisi) ‘Duncan’ grapefruit (C. paradisi)	280 280 280
Bs2	Pepper (Capsicum annuum)	Lemon (C. limon)	281
Bs2	Pepper (C. chacoense)	‘Pineapple’ sweet orange (C. sinensis)	282
AtNPR1	Arabidopsis thaliana	`Hamlin’ Sweet Orange(C. sinensis)	283,284
Stx 1A	Flesh fly (Sarcophaga peregrina)	Sweet orange (C. sinensis)	285,286
terf1	Tomato	‘Succari’ sweet orange (C. sinensis)	287,288
hrpN	Erwinia amylovora	`Hamlin’ Sweet Orange(C. sinensis)	289
CalS1	C. reshni	Lemon (C. limon)	290
pthA-nls pthA-nls	Xanthomonas axonopodis pv. citri X. axonopodis pv. citri	‘Succari’ Sweet Orange(C. sinensis) ‘Bingtang’ Sweet Orange(C. sinensis)	291 291
CtNH1	Pummelo (C. maxima)	‘Duncan’ grapefruit (C. paradisi)	292
CsMAPK1	Sweet orange(C. sinensis)	Troyer citrange	293
MdSPDS1	Apple (Malus sylvestris)	‘Anliucheng’ sweet orange (C. sinensis)	294,295
Der	Phyllomedusa sauvagii	‘Pineapple’ sweet orange (C. sinensis)	296
CsMAF1	Sweet Orange(C. sinensis)	`Hamlin’ Sweet Orange(C. sinensis)	297
RpfF RpfF	Xylella fastidiosa X. fastidiosa	‘Pineapple’ sweet orange (C. sinensis) ‘Carrizo’ citrange (C. sinensis × P. trifoliata)	298 298
hsr203 J	Tobacco (Nicotiana tabacum cv. Xanthi)	‘Jincheng’ orange (C. sinensis)	299
PPP1	Tobacco (N. tabacum cv. Xanthi)	‘Jincheng’ orange (C. sinensis)	299
gst1	Potato (Solanum tuberosum)	‘Jincheng’ orange (C. sinensis)	299
Mthionin	Synthetic modified thionin	‘Carrizo’ citrange (C. sinensis × P. trifoliata)	300
CuSTS3-1	‘Miyagawawase’ satsuma mandarin (C. unshiu)	`Hamlin’ Sweet Orange(C. sinensis)	301
CsLOB1	Sweet Orange (C. sinensis)	‘Duncan’ grapefruit (C. paradisi)	302,303

Figure 6.

Hypothetical model of the potential polyamines- (PAs-) based response and its integration with different phytohormone pathways in citrus response to citrus canker disease caused by Xanthomonas citri subsp. citri (synonym: X. axonopodis pv. citri). Solid lines with arrowhead signify positive reaction, dashed lines with blunt-heads indicate negative interaction, while round-dotted lines with arrows represent hypothetical mechanisms or uncharacterized elements. For more details and abbreviations, see the main text.

Furthermore, the introduction of antibacterial peptide sarcotoxin IA (Stx 1A) from flesh fly (Sarcophaga peregrina) into ‘Pera’ Sweet orange (C. sinensis) enhanced the canker resistance in the transgenic lines.²⁸⁵ Interestingly, the Stx 1A transgenic lines had higher levels of putrescine and its precursor _L-arginine than non-transgenic plants (wild type) under normal growth condition.³⁰⁷ Likewise, the putrescine level has been elevated in non-transgenic ‘Pera’ Sweet orangeleaves upon Xac infection,³⁰⁷ which might play a role in citrus response to Xac. Recently, the exogenous treatment of cancer-infected sweet orange plants with spermidine enhanced the endogenous content of putrescine and spermidine, but not spermine.³⁰⁸ Additionally, the exogenously applied spermidine reduced disease incidence, index, and symptom, whereas it up-regulated the transcript levels of four defense-related genes, including AOS, PR4A,glutathione peroxidase(GPX), and chitinase.³⁰⁸ Collectively, we suggest that PAs might play a key role in citrus defensive responses against citrus canker via regulation of defense-related genes and ROS homeostasis.

Additionally, PAs might be integrated with phytohormones and their metabolism-related genes in citrus response to Xcc. For instance, the transcriptomic analysis of citrus cDNA microarray showed that about 92 phytohormone metabolism-related genes were differentially regulated during the citrus response to Xcc.³⁰⁹ These genes involved in the metabolism ofabscisic acid (ABA),auxins, brassinosteroids (BRs), cytokinins (CKs), ethylene (ET), gibberellins (GAs), jasmonic acid (JA), and salicylic acid (SA), as well as the nitric oxide pathway(NO).³⁰⁹Interestingly, the Xcc-infection induced the accumulation of putrescine and spermidine,but not spermine, at 12- and 24-hours post-inoculation (hpi). The accumulation of putrescine and spermidine was associated with higher levels of JA but not the other phytohormone groups.³⁰⁹ Together,these findings suggest that the JA activation could involve a type of interaction-dependent regulation of free PAs levels, particularly putrescine and spermidine.

Furthermore, the transcriptomic analysis of putative PAs metabolism-related genes in citrus cDNA microarray showed that a putative ornithine decarboxylase (ODC) was down-regulated, while the arginine decarboxylase 1 (ADC) up-regulated after the infection with Xcc.³⁰⁹ These findings suggest that the accumulation of putrescine, and subsequently spermidine, might be due to the activation of PAs biosynthesis via thearginine-agmatine-putrescine route involving ADC, rather than the arginine-ornithine-putrescine route involving ODC. In addition, Xcc-infection induced the transcriptional levels of three polyamine oxidase genes (PAOs).³⁰⁹ PAOs are involved in the back conversion of spermidine and spermine to putrescine, which could partially explain the accumulation of putrescine after Xcc infection. Likewise, the transcriptional levels of a putative copper amine oxidase family protein (CuAO) were induced in the Xcc-infected citrus plants.³⁰⁹ Interestingly, the CuAO was up-regulated in the canker resistant ‘Meiwa’ kumquat (Fortunella crassifolia) and the moderate resistant M-9 cybrid (Cybrid-K type) after Xcc infection.³¹⁰ On the other hand, under the same conditions, CuAO was down-regulated in the canker susceptible ‘Marsh’ grapefruit (C. paradisi) plants and M-102 cybrid (Cybrid-G type).³¹⁰ Collectively, these findings suggest that CuAO might play a key role in citrus responses against Xcc infection.

PAs-based defense responses against citrus variegated chlorosis (CVC)

Citrus Variegated Chlorosis (CVC) is a destructive bacterial disease in citrus. CVC is caused by the Gram-negative, xylem-limited, insect-transmitted Gammaproteobacterium Xylella fastidiosa.^311,312 Although CVC could be transmitted by grafting, however, it is mainly transmitted by several xylem sap-sucking insects including sharpshooters (Cicadellidae), treehoppers (Membracidae), and spittlebugs (Cercopidae).In Brazil, CVC might be transmitted by 11 species of sharpshooters,³¹³ however, only two of them were found in Florida including the glassy-winged sharpshooter, Homalodisca vitripennis, and the blue-winged sharpshooter, Oncometopia nigricans,³¹⁴ but the CVC disease is not reported in the USA yet. Since X. fastidiosa colonizes the xylem vessels, it might form a biofilm that blocks the flow of xylem sap, causing a water deficit-like symptom in the upper part of the infected trees.³¹⁵ X. fastidiosa can infect most of the citrus species and hybrids causing different degrees of symptoms severity. Sweet orange (C. sinensis) varieties are highly susceptible to X. fastidiosa infection, while Rangpur lime, lemon, citron, and pummelo are considered CVC-tolerant varieties and remain symptomless, regardless of the inoculum pressure. Grapefruit, mandarins, mandarin hybrids, lemons, limes, kumquat, and trifoliate orange are moderately susceptible, showing less severe symptoms.

Upon the infection with X. fastidiosa, citrus plants develop a complex multilayered- defensive responses at the physiological, genetic, and metabolic levels.^315–317 Due to the difficulties found in the study of X. fastidiosa-citrus pathosystem, very few studies were carried out the elucidate the metabolic changes that happen in citrus plants upon the infection with X. fastidiosa. These metabolic changes included some phytohormones³¹⁸ and nitrogen metabolism.³¹⁶ Nitrogen metabolism is a key factor that affects the growth of X. fastidiosa,³¹⁹ as well as, X. fastidiosa disturbs the nitrogen metabolism of citrus plants.^315,316,320 For example, the PAs-associated amino acids, _L-proline, and _L-arginine were increased atthe initial stages of CVC development in the leaves of ‘Pêra’ sweet orange (C. sinensis) grafted on Rangpur lime (C. limonia).³²⁰ Furthermore, the diamine putrescine was detected in higher levels in leaves and petioles of X. fastidiosa-infected plants.³¹⁶ The accumulation of putrescine might be due to the induction of its precursor arginine which increased in leaves and xylem sap of X. fastidiosa-infected plants.³¹⁶ Nevertheless, it is not clear yet if the accumulation of both putrescine and arginine is a direct plant response to the X. fastidiosa infection or it is due to the water deficit-like symptom imposed by the xylem blocking since both water deficiency and pathogens attack are known to induce the endogenous PAs levels in plants. However, more investigations are required to clarify the putative PAs-based defensive response against CVC and its integration with other metabolic responses, particularly phytohormones.

Roles of PAs in the Huanglongbing (HLB; citrus greening) pathosystem

Huanglongbing (HLB; aka citrus greening disease) is a destructive bacterial disease of rutaceous plants that can infect most, if not all, citrus cultivars, hybrids, and some citrus relatives. Although Koch’s postulates of HLB have not been fulfilled yet due to the difficulty in culturing the putative bacterium, several pieces of evidence indicate that HLB is associated with a plant fastidious, phloem-limited, phytopathogenic, alphaproteobacterium Candidatus Liberibacter spp..^54,321–323 Based on the geographical distribution and the characteristic 16 S rDNA sequence, three bacterial species have been proposed to be associated with HLB.^54,324–326 ‘Ca. L. asiaticus’in Arabian Peninsula, Africa, Asia, and the Americas, Ca. L. americanusspreading only in Brazil, andCa. L. africanusinnumerous countries in Africa.^54,323–328 Among the three Ca. Liberibacter species, ‘Ca. L. asiaticus’is the dominant species and is causing huge economic losses to citrus production worldwide.^54,323The tree-to-tree transmission of Ca. Liberibacter could be by graft inoculation, and they are mainly transmitted by insect vectors called citrus psyllids.³²⁹ HLB is transmitted by two insect vectors; the African citrus psyllid, Trioza erytreae Del Guercio (Hemiptera: Triozidae), which transmits Ca. L. africanus^323,330 and the Asian citrus psyllid, Diaphorina citriKuwayama(Hemiptera:Liviidae), which transmits both ‘Ca. L. asiaticus’ and Ca. L. americanus.^323,331,332 Globally, D. citri, a phloem sap-sucking insect, is one of the most serious pests of citrus, particularly when ‘Ca. L. asiaticus’ is present³²⁹ because it harms the new leaves and fruits in addition to its ability to transmit ‘Ca. L. asiaticus’.

Both ‘Ca. L. asiaticus’ and its vector manipulate the host metabolism for their benefit to meet their nutritional needs or to neutralize the host defense responses. In addition, upon ‘Ca. L. asiaticus’-infection and/or D. citri-infestation, citrus plants develop multiple defense mechanisms to protect themselves. Previous studies showed that both ‘Ca. L. asiaticus’-infection and D. citri-infestation can cause many biochemical and metabolic alterations in citrus plants. For instance, both ‘Ca. L. asiaticus’-infection and D. citri-infestation can alter the volatile organic compounds,^{203,333–336} amino acids,^{1–6,337,338} organic acids,^1,3,6,337 fatty acids,^2,6 phytohormones,^7,8 polyamines,^4,12 carotenoids and leaf pigments^202,339 and other secondary metabolites^333,337 such as phenolic compounds^{333,340–342} and flavonoids.^{339,342–344} However, it remains unclear whether these changes benefit ‘Ca. L. asiaticus’and its vector, or if they are a plant response or both. Although many studies have been carried out to understand the metabolic-based defense responses of citrus plants against ‘Ca. L. asiaticus’ and its vector, only a few studies about the role of PAs pathway in regulating citrus defense responses and HLB symptoms development have been reported.

Based on comparative genomic studies of uncultured strains of ‘Ca. L. asiaticus’, ‘Ca. L. solanacearum’, and their closest cultured species L. crescens BT-1, ‘Ca. L. asiaticus’ was found to lack the biosynthetic pathways and transporters for some amino acids, as well as, PAs biosynthesis pathway.³⁴⁵ For instance, ‘Ca. L. asiaticus’ lack the de novo biosynthetic pathways for six amino acids including _L-cysteine, _L-histidine, _L-phenylalanine, _L-proline, _L-tryptophan, and _L-tyrosine which were reported to be produced by L. crescens.^345–347 Likewise, the putrescine biosynthetic pathway was found in L. crescens but it was reported to be partially incomplete in ‘Ca. L. asiaticus’, which lacks the ODC enzyme.³⁴⁵ In other words, ‘Ca. L. asiaticus’ can metabolize the putrescine but it lacks the necessary enzymes for its de novo biosynthesis. Therefore, ‘Ca. L. asiaticus’ might depend on their host plant to get its PAs needs. Recently, we showed that ‘Ca. L. asiaticus’ infection induces the total PAs, ornithine, and putrescine content,¹² which may establish a vital host-dependence for its nutrition needs. However,‘Ca. L. asiaticus’ genome does not possess any PAs transporters that could import putrescine or its precursor, ornithine, based on the previous comparative genomic study with L. crescens³⁴⁵ or in-depth prediction of ABC transporters in ‘Ca. L. asiaticus’.³⁴⁸

Herein, we are going to discuss the previous and most recent integrative studies using transcriptomics, proteomics, and metabolomics approaches to understand the citrus response(s) against ‘Ca. L. asiaticus’ and/or its insect vector D. citri. We believe that the metabolic responses of citrus plants may result from host cellular functions for defense reactions, or it may result from the manipulation of metabolic pathways by ‘Ca. L. asiaticus’ and/or D. citri for their benefits to fulfill their nutritional needs, or to induce symptoms development.

Deciphering the Role of Citrus-PAs in HLB Symptoms Development

The most characteristic symptom of HLB is the blotchy mottle, which is an asymmetrical discoloration/chlorosis across the mid-vein of the leaf with patches of yellow and green islands.^323,349 Symptoms of HLB are induced by both the pathogen and its vector, due to alteration in many physiological aspects such as phytohormonal levels, carbohydrate status,⁷ and carotenoids and leaf pigments content.^202,339 Additionally, PAs might be involved in HLB symptom development via the production of H₂O₂. Briefly, diamine oxidase (DAO) and PAO catalyze the oxidation of putrescine and spermidine, respectively, to produce γ-aminobutyric acid (GABA) and resulting in accumulation of H₂O₂ within the apoplast.^304,305 The accumulation of H₂O₂ in citrus plants after the infection with ‘Ca. L. asiaticus’ has been reported previously.³⁵⁰ In our previous study, GABA was accumulated and both D-amino acid oxidase PA4548 (aka diamine oxidase CsDAO) and CsPAO were expressed at higher levels in ‘Ca. L. asiaticus’ -infected plants,¹² which is in agreement with this common feature of pathogen-host interaction in many previous studies.^304,305 Thus, we hypothesize that the induction of PAs and their catabolic genes (CsDAO and CsPAO) are correlated with the accumulation of GABA and H₂O₂ which are working as a response to HLB to create an incompatible interaction. Nevertheless, the alteration in DAO and PAO, and the accumulation of H₂O₂ have to occur in the appropriate location of the plant to be effective in the creation of incompatible interactions between the host and the pathogen.³⁰⁵ Interestingly, ‘Ca. L. asiaticus’ could survive the toxic effects of accumulated H₂O₂ using its own peroxidase.³⁵⁰ The detoxification system of citrus plants, however, might not be sufficient to reduce the high H₂O₂levels, which may eventually become toxic to the leaf tissue and cause the characteristic blotchy mottle symptoms that appear after ‘Ca. L. asiaticus’-infection.³⁵⁰

PAs-based defense responses against HLB

Recently, we showed that ‘Ca. L. asiaticus’-infection induced the PAs biosynthesis, leading to an accumulation of ornithine and putrescine (Figure 7), and their precursor _L-arginine, which detected only in ‘Ca. L. asiaticus’-infected plants, but not other treatments.¹² Similar putrescine induction was found in the juice from ‘Valencia’ Sweet Orange (C. sinensis) fruits from ‘Ca. L. asiaticus’-infected trees³³³ and leaves from D. citri-infested trees⁴ (Table 10). Furthermore, the semi-tolerant varieties such as ‘Volkamer’ lemon (C. volkameriana), Alemow (C. macrophylla),’Mexican’ lime (C. aurantifolia), and ‘Palestine Sweet’ Lime (C. aurantifolia); and the very tolerant variety, Swingle(C. latipes)³⁵⁶ has a greater putrescine content. Moreover, the putrescine conjugate, feruloyl putrescine, was found at high levels in juice from HLB-symptomatic ‘Valencia’ sweet orange (C. sinensis) fruits but did not change significantly in both ‘Hamlin’ or ‘Midsweet’ oranges compared to juice from HLB-asymptomatic and healthy fruit.³³³ Collectively, these findings indicate that putrescine plays a key role in citrus response/tolerance against ‘Ca. L. asiaticus’. In plants, putrescine is synthesized through two different routes; from arginine using ADC or from ornithine using ODC.³⁶⁰ Interestingly, both CsADC and CsODC were expressed at higher levels in ‘Ca. L. asiaticus’-infected plants indicating that both putrescine biosynthetic pathways are jointly working to enhance the PAs content (Figure 8). Likewise, the arginine pathway metabolites (ornithine, citrulline, and _L-proline) were found in much higher abundance in the HLB-susceptible cultivar in response to ‘Ca. L. asiaticus’-infection³⁵³

Figure 7.

Representative chromatogram of detected non-proteinogenic amino acids (NPAAs) and polyamines (PAs) compoundsof Valencia sweet orange (C. sinensis) leaves after derivatization with MCF using GC-MS running in full-scan mode. Citrus leaf NPAAs and PAs compounds were extracted from frozen tissue using methanol 80% containing 0.1% HCl 6N as described in our previous studies.^6,351 Leaf extract was derivatized with methyl chloroformate (MCF) following the protocol of Hijaz and Killiny 2014,²⁰¹ with slight modifications as described by Killiny and Nehela 2017.⁶ Derivatized samples were analyzed using a Clarus 680 GC-MS system (Perkin Elmer, Waltham, MA, USA) fitted with a ZB-5 MS GC column (Phenomenex, Torrance, CA, USA) using the GC thermo-program described in our previous study.⁶ Chromatographic characteristics, MS ion identification, and GC-MS chromatograms analysis were performed according to Killiny and Nehela 2017.⁶

Table 10.

Changes in the endogenous content of major polyamines (PAs), γ-Aminobutyric acid (GABA), Ornithine (Orn; a PAs precursor), and their associated amino acids in citrus and its relative species as affected by Candidatus Liberibacter asiaticus (the causal agent of Huanglongbing) and/or its vector, Diaphorina citri.

‘Cultivar’/Variety/(species)	Tissue	Treatment/experimental design	Pro a	Arg	GABA	Orn	Put	Spd	Spm	Reference
‘Valencia’ Sweet Orange (Citrus sinensis) b	Fruits	Orange juice from fruits harvested from healthy vs. ‘Ca. L. asiaticus’-infected trees in two seasons	-	-	-	-	▲	-	-	333
‘Valencia’ Sweet Orange (C. sinensis)	Leaves	Leaves from HLB‐infected, zinc‐deficient, and healthy trees using GC‐MS	▲	-	-	-	-	-	-	1
Four citrus varieties c	Leaves	Two HLB-tolerant citrus varieties vs two HLB-sensitive varieties	↑↓	-	-	-	-	-	-	2
‘Duncan’ grapefruit (C. paradisi)	Leaves	New flushes of healthy vs HLB-infected trees at 14 and 20 weeks after inoculation	▲	-	-	-	-	-	-	2
‘Valencia’ Sweet Orange (C. sinensis)	Fruits	Orange juice obtained from healthy, asymptomatic, and symptomaticfruits	▼	▼	●	-	-	-	-	3
‘Valencia’ and ‘Hamlin’ Sweet Orange (C. sinensis)	Fruits	Orange juice from healthy, HLB-asymptomatic, and -symptomatic oranges	▼	-	▲	-	●	-	-	338
‘Owari’ Satsuma orange (C. unshiu)	Leaves	Effect of D. citri infestation on leave metabolites (healthy vs D. citri-infested)	▲	▲	-	-	▲	●	●	4
‘Valencia’ Sweet Orange (C. sinensis)	Leaves & Roots	HLB-asymptomatic vs. -symptomatic tissues from 6-year-old commercial orchard	●	-	-	-	-	-	-	352
Six citrus varieties d	Leaves	Comparing five tolerant hybrids with HLB-sensitive cultivar (healthy vs ‘Ca. L. asiaticus’ -infected)	▲	▲	-	▲	-	-	-	353
Different hosts e	Phloem sap	The phloem sap composition of orange jasmine vs. Curry leaf tree using GC-MS	▲	▲	▲	▲	▲	-	-	354
Fourteen citrus varieties f	Phloem sap	13 citrus varieties with different tolerance degrees to ‘Ca. L. asiaticus’ (MCF derivatization)	▲	-	▲	-	-	-	-	5
Alemow (C. macrophylla)	Leaves	The effect of D. citri feeding (0, 5, 15–20, or 25–30 insect/plant) on leaf metabolites.	●	▲	▲	-	-	-	-	355
Fourteen citrus varieties g	Phloem sap	14 citrus varieties with different tolerance degrees to ‘Ca. L. asiaticus’ (TMS derivatization)	▲	▲	▲	-	▲	-	–	356
‘Valencia’ Sweet Orange (C. sinensis)	Leaves	Tested the effect of ‘Ca. L. asiaticus’ -infection and/or D. citri-infestation on leaf metabolites	▲	-	▲	-	-	-	-	6
Different hosts h	Phloem sap	Four ‘Ca. L. asiaticus’ -permissive host plants vs. three non-permissive Rutaceae plants	↑↓	↑↓	↑↓	-	-	-	-	357
‘Ray Ruby’ grapefruit (C. paradisi) i	Leaves	Leaves from HLB‐infected vs. healthy trees using GC‐MS	▲	▲	-	-	-	-	-	358
‘Hamlin’ Sweet Orange (C. sinensis)	Fruits	Orange juice obtained from healthy, asymptomatic, and symptomaticfruits	▲	●	-	-	-	-	-	359
‘Valencia’ Sweet Orange (C. sinensis)	Leaves	Tested the effect of ‘Ca. L. asiaticus’ -infection and/or D. citri-infestation on leaf PAs	▲	▲	▲	▲	▲	-	-	12

^a(▲) signifies for increased levels, (▼) signifies for decreased levels, (●) signifies for no significant change, (↑↓) signifies for varied responses, and (˗) signifies for not-reported compounds. Abbreviations: Pro:_L-proline, Arg:_L-Arginine, GABA:γ-Aminobutyric acid, Orn: Ornithine, Put: putrescine, Spd: spermidine, and Spm: spermine.

^bdetermined as feruloyl putrescine³³³

^cThey tested the leave metabolites of two HLB-tolerant citrus varieties (‘Carrizo’ citrange [×Citroncirus webberi] and Trifoliate Orange [P. trifoliata]) and two HLB-sensitive varieties (‘Madam Vinous’ sweet orange [C. sinensis] and ‘Duncan’ grapefruit [C. paradisi]) using GC-MS after derivatization with methoxyamine and N-methyl-N-trifluoroacetamide (MSTFA).²

^dThey tested the leave metabolites of five tolerant hybrids in comparison with HLB-sensitive cultivar (healthy vs ‘Ca. L. asiaticus’-infected) using GC-TOF MS after methoximation and trimethylsilylation derivatization. The tested varieties included: Cleopatra mandarin (C. reticulata), Carrizo citrange (C. sinensis × P. trifoliata), US-802 (C. grandis ‘Siamese pummelo’ × P. trifoliata ‘Gotha Road’), US-812 (C. reticulata ‘Sunki’ × P. trifoliata ‘Benecke’), US-897, and US-942 (C. reticulata ‘Sunki’ × P. trifoliata ‘Flying Dragon’).³⁵³

^eThe author studied the phloem sap composition of Valencia sweet orange (C. sinensis), Curry leaf tree (Bergera koenegii; a specific host for D. citri, but not ‘Ca. L. asiaticus’), and orange jasmine (Murraya paniculate; an alternative host for both D. citri and ‘Ca. L. asiaticus’).³⁵⁴

^fThey tested the phloem sap composition of 13 citrus varietieswith different tolerance degrees to ‘Ca. L. asiaticus’ using GC-MS after methyl chloroformate (MCF) derivatization. The tested varieties included: ‘Valencia’ sweet orange (C. sinensis), ‘Duncan’ grapefruit (C. paradise), ‘Ruby Red’ grapefruit (C. paradisi), ‘Madam Vinous’ sweet orange (C. sinensis), Alemow (C. macrophylla), ‘Mexican’ lime (C. aurantifolia), ‘Palestine Sweet’ Lime (C. aurantifolia), ‘Sour’ orange (C. aurantium), ‘Volkamer’ lemon (C. volkameriana), Chinese box orange (Severinia buxifolia), Poncirus trifoliata, Carrizo citrange trifoliate hybrid (‘Washington’ sweet orange [C. sinensis] × P. trifoliata), and Swingle(C. latipes).⁵

^gThey tested the metabolite signature of the phloem sap of 14 citrus varietieswith different tolerance degrees to ‘Ca. L. asiaticus’ using GC-MS after trimethylsilyl (TMS) derivatization. The tested varieties included: ‘Valencia’ sweet orange (C. sinensis), ‘Pineapple’ sweet orange (C. sinensis), ‘Duncan’ grapefruit (C. paradise), ‘Ruby Red’ grapefruit (C. paradisi), ‘Madam Vinous’ sweet orange (C. sinensis), Alemow (C. macrophylla), ‘Mexican’ lime (C. aurantifolia), ‘Palestine Sweet’ Lime (C. aurantifolia), ‘Sour’ orange (C. aurantium), ‘Volkamer’ lemon (C. volkameriana), Chinese box orange (Severinia buxifolia), Poncirus trifoliata, Carrizo citrange trifoliate hybrid (‘Washington’ sweet orange [C. sinensis] × P. trifoliata), and Swingle(C. latipes).³⁵⁶

^hThey tested the phloem sap composition of four ‘Ca. L. asiaticus’-permissive host plants vs. three non-permissive Rutaceae plants using a High-Speed Amino Acid Analyzer. The tested ‘Ca. L. asiaticus’-permissive host plants included: ‘Rio Red’ Grapefruit (C.paradisi), ‘Marrs’ Sweet orange (C. sinensis), ‘Valley’ Lemon (C. limon), and ‘Madagascar’ periwinkle (Catharanthus roseus); while the non-permissive plants included: Curry leaf (Murraya koenigii), ‘Lakeview’ Orange jasmine (Murraya exotica), and White sapote (Casimiroa edulis).³⁵⁷

ⁱThey also studied the metabolic changes in leaves from ‘Ca. L. asiaticus’‐infected vs. healthy trees of Hamlin orange (C. sinensis), Murcott mandarin (C. reticulata), but without any significant differences between healthy and ‘Ca. L. asiaticus’‐infected plants for those varieties.³⁵⁸

Figure 8.

Hypothetical model of the potential polyamines- (PAs-) based response and its integration with other metabolic pathways in citrus response to Huanglongbing (aka citrus greening disease) caused by Candidatus Liberibacter asiaticus. Solid lines with arrowhead signify positive reaction, dashed lines with blunt-heads indicate negative interaction, while round-dotted lines with arrows represent hypothetical mechanisms or uncharacterized elements. For more details and abbreviations, see the main text.

In one of our recent studies, we proposed a hypothetical model for the effect of ‘Ca. L. asiaticus’ infection and/or D. citri infestation on the proteinogenic amino acids, NPAAs, PAs, and tricarboxylic acid- (TCA-) associated compounds of Valencia sweet orange (C. sinensis) leaves.¹² In this model, we proposed that citrus plants have an alternative pathway(s) that contribute to the flux toward succinate thereby increasing the availability of fumarate and citrate rather than as an intact TCA cycle (Figure 8). GABA-shunt and PAs biosynthesis pathway are potential candidates for these alternative pathways (from α-ketoglutarate to succinate) and they might be enhanced after under specific abiotic and biotic stresses such as ‘Ca. L. asiaticus’-infection and D. citri-infestation.¹² Briefly, in our published model, we proposed that ‘Ca. L. asiaticus’-infection might accelerate the conversion of α-ketoglutarate to glutamate, then to GABA, which results in an accumulation of GABA in the cytosol. Subsequently, GABA is transported from the cytosol to the mitochondria using the GABA permease (GABP), to support the accumulation of succinate directly or indirectly. Interestingly, GABA was detected at high concentration in the HLB-tolerant citrus varieties such as P. trifoliate, Swingle(C. latipes), ‘Volkamer’ lemon (C. volkameriana), Carrizo citrange trifoliate hybrid (‘Washington’ sweet orange [C. sinensis] × P. trifoliata), and Chinese box orange (Severinia buxifolia), ^5,356 which support our point of view that GABA plays an important role in citrus response against HLB.

Another pathway that could be involved in the enrichment of the TCA cycle is the PAs biosynthesis pathway.PAs biosynthesis pathway could be involved in the enrichment of the TCA cycle directly through the production of fumarate or indirectly via enhancement of GABA-shunt. The dicarboxylic acid ‘fumarate’ is produced during the conversion of citrulline to argininosuccinate by argininosuccinate synthase (ASS), then to the amino acid arginine using argininosuccinate lyase (ASL; aka argininosuccinase) which splits argininosuccinate to release fumarate and arginine.³⁶¹ Our findings showed that both CsASS and CsASL were upregulated after ‘Ca. L. asiaticus’-infection, which resulted in the accumulation of arginine (detected only in ‘Ca. L. asiaticus’-infected plants). Together, this might enhance the TCA cycle intermediate fumarate directly (Figure 8). Similar induction of fumarate was observed during the necrotrophic phase of rice blast fungus Magnaporthe grisea³⁶² and the maize anthracnose fungus Colletotrichum graminicola.³⁶³

Although several studies have been reported the effect of ‘Ca. L. asiaticus’ infection on citrus metabolites, however, the role of PAs in the HLB pathosystem remains poorly understood because some of these studies reported the effect of ‘Ca. L. asiaticus’ infection on the endogenous content of putrescine only (Table 10) but, to our knowledge, the effect of ‘Ca. L. asiaticus’ infection on both spermidine and spermine was not reported till now. We believe that all PAs, not onlyputrescine, might play a role in citrus defense against ‘Ca. L. asiaticus’ and its vector, via the integration with phytohormones, phenolic amines (such as synephrine), GABA shunt, and stress-associated amino acids, such as proline (Figure 8).

The phenolic monoamine ‘synephrine’ and its precursor, octopamine, are compounds that might be involved with PAs in the defense response against HLB. Synephrine was reported previously in several citrus species and cultivars.^364,365 In citrus, synephrine is biosynthesized from tyrosine via a pathway involving tyramine, N-methyltyramine, then to synephrine using tyrosine decarboxylase (TDC), tyramine N-methyltransferase (TNMT), and an unidentified N-methyl-tyramine-β-hydroxylase (NMTβH), respectively.^365–367 While it synthesized through tyramine, octopamine, then to synephrine in animals, using TDC, TβH and PNMT, respectively.³⁶⁶ Although octopamine is not an important intermediate in plants as in animals, our results showed that the octopamine level was elevated in the presence of ‘Ca. L. asiaticus’.Furthermore, the transcript level of CsTDC was increasedin all studied treatments. Because only TDC has been isolated and cloned from plants, while other enzymes have not yet been cloned,³⁶⁷ our image about the synephrine-PAs pathway is not fully understood and it might require further studies.

Although many previous studies had investigated the _L-proline content after ‘Ca. L. asiaticus’-infection and/or D. citri-infestation (Table 10), however, the integration of _L-proline in the PAs-based defense responses of citrus plants against HLB is poorly studied. Higher levels of the amino acids _L-proline, in healthy plants were characteristic of the most sensitive variety ‘Madam Vinous’ sweet orange (C. sinensis) when compared to HLB-tolerant varieties such as ‘Carrizo’ orange (Carrizo citrange) and P. trifoliate.² Likewise, ‘Ca. L. asiaticus’-permissive hosts such as ‘Rio Red’ Grapefruit (C.paradisi), ‘Marrs’ Sweet orange (C. sinensis), ‘Valley’ Lemon (C. limon), and ‘Madagascar’ periwinkle (Catharanthus roseus) had a greater content of _L-proline compared to the non-permissive plants such as Curry leaf (Murraya koenigii), ‘Lakeview’ orange jasmine (Murraya exotica), and white sapote (Casimiroa edulis).³⁵⁷ On the other hand, our previous study showed that the phloem sap of orange jasmine (M. paniculate; an alternative host for both D. citri and ‘Ca. L. asiaticus’) had a higher concentration of _L-proline compared with the HLB-sensitive variety ‘Valencia’ sweet orange (C. sinensis) and the specific host of D. citri, Curry leaf tree (Bergera koenegii).³⁵⁴ Furthermore, the phloem sap of the semi-tolerant varieties such as’Sour’ orange (C. aurantium) and ‘Volkamer’ lemon (C. volkameriana); tolerant varieties such as Carrizo citrange trifoliate hybrid (‘Washington’ sweet orange [C. sinensis] × P. trifoliata) and Chinese box orange (Severinia buxifolia); and the very tolerant variety Swingle(C. latipes) had a higher _L-proline content compared to HLB-sensitive varieties.^5,356

In addition, higher levels of_L-proline have been previously reported in symptomatic ‘Ca. L. asiaticus’-infected^1,2,6,12,358 and D. citri-infested leaves.^4,355 However, _L-proline was reported to be reduced in the juice of symptomatic HLB-infected ‘Valencia’ and ‘Hamlin’ Sweet Orange (C. sinensis) fruits.^3,338 In disagreement, _L-proline was accumulated in the juice obtained from asymptomatic and symptomatic HLB-infected ‘Hamlin’ Sweet Orange (C. sinensis) fruits.³⁵⁹ Nevertheless, in both studies, the PAs precursor amino acid,_L-arginine, was not changed or slightly reduced in juice from HLB-symptomatic fruits.^3,359 Interestingly, _L-arginine was found at higher levels in the phloem sap of the alternative host, orange jasmine (M. paniculate)³⁵⁴ and the semi-tolerant varieties such as ‘Palestine Sweet’ Lime (C. aurantifolia) and ‘Volkamer’ lemon (C. volkameriana).³⁵⁶ Moreover, _L-arginine was accumulated in citrus leaves after the infection with ‘Ca. L. asiaticus’^12,358 and the infestation with D. citri.^4,12,355

In addition, PAs might be integrated with phytohormones to mediate the citrus responses to ‘Ca. L. asiaticus’ and its vector, D. citri(Figure 8). Previously, we showed that multiple phytohormonal signaling mediates citrus responses to the HLB disease.⁸ Briefly, ‘Ca. L. asiaticus’-infection and D. citri-infestation alert the auxins, SAs, JAs, and ABA profiles of citrus leaves, but do not affect both cytokinins, and gibberellins levels.⁸ Likewise, the transcriptional analysis of sweet orange trees showed that the solo infection with ‘Ca. L. asiaticus’-B232 disrupted the ABA, auxin, ethylene, JA, and SA, but did not affect the brassinosteroid, cytokinin, or gibberellins pathways.^238,239 Interestingly, the accumulation of different phytohormone groups and their related pathways in ‘Ca. L. asiaticus’-infected plants wereassociated with high levels of putrescine and ornithine.¹² Furthermore, the exogenous application of GABA increased the level of different phytohormones in ‘Valencia’ sweet orange (C. sinensis).³⁶⁸ Likewise, the exogenous putrescine application (300 mg L⁻¹) increased the IAA, SA, ABA and endogenous putrescine content, while it decreased the gibberellins content of sour orange (C. aurantium) seedlings under salt stress.369 Collectively, these findings support our suggestion that PAs might be integrated with phytohormones to mediate the citrus responses to ‘Ca. L. asiaticus’ and its vector, D. citri, however, more investigations are required for better understanding of PAs-based defense responses against HLB.

Conclusion Remarks and Future Prospects

In this review article, we comprehensively discussed the published literature on the role of PAs in citrus response to abiotic and biotic stress. We explored the PAs-based defensive responses of citrus against abiotic stress such as low- and high-temperature, salinity, heavy metals, drought stress. As well as, we deciphered the potential roles of PAs in citrus response to biotic stressors including virus and viroid infection, fungal and oomycetes infection, and bacterial infection, particularly, the vector-borne phytopathogenic bacteria.

In conclusion, PAs play a key role in protecting the citrus plant from stress-associated damages and enhancing stress tolerance through a multi-layered defense system. The PAs-based defense system involving; 1) the modulation of cellular PA levels, 2) interaction with other metabolic pathways, particularly phytohormones such as salicylic acid, auxins, jasmonic acid, abscisic acid, ethylene, cytokinins, and brassinosteroids, 3) activation of glutamine-Proline pathway and GABA Shunt, 4) activation the cascades of CBF/COR genes and SOS pathway members (SOS3, SOS2, and SOS1), and/or 5) the interaction with other defensive mechanisms such as Ca²⁺ influx, ROS detoxification, and induction of PAL enzymes and microRNAs.

We believe that the knowledge gained so far about the roles of PAs in citrus responses underpin our understanding PAs-based defensive responses and provide valuable information for the potential use of PAs or their synthetic analogs a promising approach to practical applications to enhance stress tolerance in citrus plants.

Although tremendous efforts have been done to understand the role of PAs in citrus responses to various stressors, however, most of them have been relatively simple and similar. Most previous studies have focused on the alteration of endogenous PAs content under biotic and abiotic conditions, modulation of cellular PAs content via exogenous PAs supplementation, or the induction of endogenous PAs production via genetic manipulation of PAs biosynthetic genes. Nevertheless, very few previous studies have been made to identify, clone, and functionally characterize PAs biosynthetic genes from citrus and its relative species including CsSAMDC⁷⁹ and PtADC.⁸² Therefore, further studies are required for cloning, biochemical identification, function characterization, and expression analysis of the genes involved in PAs biosynthesis pathway in citrus, as well as its phytopathogenic microorganisms. For example, the phytopathogenic bacterium ‘Ca. L. asiaticus’ (the causal agent of citrus greening disease) has incomplete PAs biosynthesis pathway, which lacks ODC.³⁴⁵ Thus, ‘Ca. L. asiaticus’ is mainly depending on their host for its nutrition needs of PAs and some other metabolites, which hinders culturing efforts. Therefore, we believe that addition PAs to the ‘Ca. L. asiaticus’ growth medium may improve growth.

Nevertheless, there are still numerous unanswered questions in this exciting area of research, which could be viewed as a promising new area of research and as an inspiring topic for further progress in this field.Firstly, although numerous observations have been addressed the PAs accumulation in response to both biotic and abiotic stresses, however, the molecular mechanisms behind the PAs accumulation are still largely unknown. Thus, further investigations are required for a better understanding ofhow the PAs biosynthetic and catabolic genes are regulated and the regulation network associated with those genes at the transcriptional and post-transcriptional levels. Secondly, the cellular translocation of PAs is poorly studied. Therefore, further studies are required to explore the PAs production sites, the PAs cellular localization, and their transporters within plant cells.Thirdly, although several hypothetical models have been proposed previously to explain the potential mode of action of PAs and their metabolites during the stressful conditions, however, the molecular mechanisms behind their roles in stress tolerance should be clearly deciphered. Most of these previous models focused on the involvement of PAs in the activation of antioxidant enzymes for ROS detoxification. In this review, we discussed the involvement of PAs with different phytohormone groups in citrus response.Nevertheless, further physiological, metabolic, transcriptomic, proteomic and molecular research is required to uncover the exact mode of action of PAs in citrusresponse and to decipher the PAs-phytohormones interactions and the metabolic cross-talk between them during biotic and abiotic stress. Last but not least, our knowledge about the signaling cascades that linking the PAs biosynthetic genes and stress responses in citrus plants is still very limited and very far from being well defined. In other plant species, the transcription factors regulating the ADC genes have been identified previously, while those whose regulate other PAs biosynthetic genes remain unknown. Therefore, deciphering the signaling cascades of PAs in citrus plants under both normal and stressful conditionsis a necessity and could be a new exciting area of research in citrus physiology.

Funding Statement

This work was generously supported by grant No. 2016-70016-24844 for NK from the National Institute of Food and Agriculture (NIFA-USDA).

Author Contributions

N.K., together with Y.N., conceptualized the idea. Y.N. and N.K. contributed to digging into literature and finalizing the figures. Both authors worked together in drafting the manuscript. Finally, N.K. and Y.N. revised and finalized the manuscript. All authors have read and agreed to the published version of the manuscript.

Disclosure statement

The authors declare there is no conflict of interest.